TWI284605B - Automatic train operating device - Google Patents

Abstract

To put the learning of properties forward automatically also after business travel by reducing the time and labor required for tuning. This on-line processes data acquired during travel of a train and automatically learns train properties during travel of the train, with the control parameters of train travel, the physical properties of the train itself during train operation, and the physical values showing the rout properties as the train properties, based on the data obtained hereby and the data obtained beforehand, and this performs the automatic operation of the train, using the train properties obtained by these automatic properties.

Description

1284605 (1) EMBODIMENT DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an automatic train running device that does not require a driver to stop an automatic operation of a train at a specific time at a specific time. [Prior Art] Automatic train running device (hereinafter referred to as "ΑΤΟ") is automatic

The train is operated between stations, and the train is stopped at a specific parking position at a specific time. Fig. 47 is a diagram showing an example of the system configuration of the electric train having such a crucible. The automatic train control device (ATC) not shown on the figure inputs a speed limit signal to the automatic train running device 1, and the data bank 3 inputs the route conditions such as the slope and curvature, the vehicle condition, and the running time table to the automatic train running device 1. And stored information such as driving resistance. Further, the automatic train running device 1 calculates the current vehicle position based on the vehicle position detected by the ground sub-detector 10 and the vehicle speed detected by the φ speed detector 9, and inputs a thrust command F cmd to the drive brake device 2, Indicates the thrust that should be provided at that point in time. At this time, the thrust command F cmd in this manual is defined as both the traction command for vehicle acceleration and the braking force command for vehicle deceleration. The traction command is the thrust command F cmd > 0, and the thrust force is the thrust command F cmd < 0. The drive brake device 2 is composed of a VVVF (variable voltage, variable frequency) variable frequency transformer inverter 4, a main motor 5, a brake control device 6, and a mechanical brake 8. The main motor 5 and the wheel 7 running on the track 11 are implemented

-5- 1284605 (2) Mechanical connection, the arrangement of the mechanical brake 8 is the same as the mechanical 煞 from the thrust command F cmd to the actual thrust of the wheel 7 'when the traction force is obtained and the braking force is obtained Therefore, the explanations are as follows. When the traction is obtained, the thrust command F cmd ( > 0 ) is input to the variable frequency inverter 4 . The variable frequency variable voltage inverter 4 controls the torque of the main motor 5

In order to get the traction force with the thrust command F cmd. At this time, the brake control device 6 and the mechanical brake 8 do not perform an operation. When the braking force is obtained, the thrust command F cmd (<0) is input to the brake control device 6 instead of the inverter variable voltage inverter 4. First, the brake control device 6 outputs a thrust command, that is, a brake force command, to the variable frequency inverter 4 . The inverter variable voltage inverter 4 feeds back the electric vehicle force F elec outputted via the main motor 5 to the brake control device 6. In order to obtain the thrust command F cmd, that is, the braking force of the braking force command, the brake control device 6 firstly activates the electric vehicle force F elec and compensates for the shortage of the electric power with the mechanical braking force F me ch of the mechanical brake 8 . Part of the way to control the mechanical brakes 8. Therefore, the mechanical braking force F mech is as follows. F mech = F cmd - F elec (1) As shown in Fig. 48, the automatic train running device 1 includes a tentative driving plan unit 1, a best driving plan unit 13 and a thrust command generating unit 14. The Provisional Driving Plan 12 will generate a tentative driving mode (F0 (X), V0 (X)) as the initial 为 for the purpose of producing the best driving mode. At this time, the driving mode indicates the thrust Fn (X) and the speed Vn (X) of the position X on the route in a manner corresponding to a series of positions. The best driving plan department 13 will be based on the -6- 1284605. (3) According to the tentative driving mode (F0 (x), V〇 (x)) and the storage of the database 3, the best driving mode F1 of the train is planned ( x ). In the optimal driving mode FI (X) generated, the thrust command generating unit 14 outputs a thrust command Fcmd to the variable frequency transformer inverter 4 according to the detected position of the train, the detection speed, and the speed limit signal of the ATC. The thrust at this point should be output.

When planning the best driving mode for a train, there are generally a number of possible driving modes. In particular, unlike the morning and evening when the timetable is too long, the number of trains running on the train is small during the day, morning, or late at night, because the train runs at a longer interval, so the plan has a larger margin, and the driving plan is There are fewer restrictions. Japanese Laid-Open Patent Publication No. Hei 8-216885 and Japanese Patent Laid-Open No. Hei 5-193502 disclose the best driving plan for energy saving as an evaluation item. However, the energy savings of these known examples are not taken into consideration from the standpoint of energy loss caused by the driving/braking control of trains such as the drive unit and the brake unit. In contrast, the "Basic Review of the Effective Use of Recycling Energy by Changing the Brake Mode" (The 7th Annual Meeting of the Japan Railway Technology Joint Conference) and the "Review of the Practicalization of Pure Electric Vehicles" (National Electrical Society National Convention 5 - In 244), the braking mode of the train, especially the driving mode of the energy loss caused by the mechanical braking caused by the braking, is reviewed. However, the energy loss caused by the drive brake control of the train is also generated during the drive control. In addition, when the brake is controlled, there are other factors that cause energy loss in addition to the mechanical brake. Therefore, the minimization of integrated energy loss cannot be achieved. 1284605 (4) / Problem to be Solved by the Invention The object of the present invention is to comprehensively evaluate the energy loss caused by train driving brake control, to minimize the energy loss between stations, and to realize energy-saving driving. Therefore, the following is a brief description of the energy loss of the present invention.

The damage caused by train driving will change due to the driving mode, and the machines that may cause the loss of the car are mainly divided into the following two categories. The first is the energy loss of the power inverter of the inverter variable frequency inverter 4 and the main motor 5. These losses can be expressed as a function of thrust and speed. The second is the energy loss caused by the mechanical brakes. Observing the acceleration and deceleration of the train from the point of view of energy flow, and ignoring the energy loss and driving resistance of the above-mentioned electric machine, during the running acceleration, the inverter variable frequency inverter 4 and the generator motor are transmitted via the unillustrated overhead line. The electric energy provided by the 5th drive device is converted into the kinetic energy of the vehicle, and in the φ deceleration of the electric vehicle, the kinetic energy of the vehicle is converted into electric energy and the power is generated again. In this ideal state, there is no energy loss. However, in the deceleration of the electric vehicle, when the braking force command of the electric vehicle or the driver exceeds the braking force that can be output by the electric machine, the mechanical braking device 8 compensates for the insufficient braking force, and the deceleration is maintained at a specific speed. When the mechanical brake 8 performs this action, the kinetic energy of the vehicle is consumed in a thermal manner, which is energy loss. In the present invention, the portion of the loss caused by the execution of the mechanical brake is defined as the vehicle loss. This brake loss occurs when the brake force command exceeds the allowable amount of the electric machine, that is, the drive load -8 - 1284605 (5), and the load on the power supply side does not exist in accordance with the regenerative power. In the latter case, if the drive unit obtains the braking force command, it will control %.

The variable frequency variable voltage inverter 4 causes the main motor 5 to output a matching braking force. At this time, the kinetic energy of the vehicle is converted into the regenerative energy of the power source. However, if there is no load corresponding to the regenerative power on the power supply side, that is, if there is no train in the acceleration, excess regenerative power is generated, thus causing the wiring. The voltage rises. Therefore, in order to suppress the rise of the overhead voltage, the drive device performs control for suppressing the braking force. This is called light load regeneration control. In this light load regeneration control operation, the main motor 5 outputs a braking force smaller than the braking force command. At this time, the lack of power will be compensated by the power of the mechanical brakes. When implementing energy-saving operation, it is important to plan an appropriate driving mode plan and actually implement driving according to the driving mode. A means for realizing the operation in accordance with the driving mode, and an automatic train running device (ΑΤΟ) and an automatic train stopping device (TASC), which are automatically generated by the driver without a driver, are well known. With these devices, it is possible to smoothly push the actual thrust to achieve the best driving mode. However, the system is complicated and costly because it directly targets the driving brake device of the vehicle and requires the ground device for the purpose of position detection. On the other hand, by using the thrust of the driver to indicate the best plan, it is possible to expect a train that is close to the planned driving mode by the skill of the driver. This is the operation support device. When such an operation support device is used, the energy saving effect is preferable to the use of the ΑΤΟ and TASC because of the driver's reaction delay, etc., however, it is only necessary to perform an instruction to the driver.

Cs' -9- (6) (6)

1284605 has no direct relationship with the driving brake device of the vehicle, so it has advantages and advantages. In addition, since it is the operation of the driver, the ground-based equipment for the purpose of position detection is turned on. Take advantage of this cost and get better cost-effective. Moreover, in recent years, the driver's driving technique has been degraded, so that it is necessary to adjust the problem of the reduction in driving skills at any time according to the judgment of the driver. Further, the automatic train running device has been put into practical use as a speed that can be chased, and a certain degree of margin with a limited speed. However, there are quite a few places where the characteristics of the vehicle and the route are controlled by the error of the ΡΙ control, etc., and the operation and control of the parameters and the control parameters are adjusted in the current situation and each route. In addition, it is also possible to formulate a driving plan and implement a train running device based on it. A simple train driving model when planning a driving plan. The simplest is the way in which the trains of its objects can be represented. F - Fr = Μ · a ... ( 7 ) At this time, the F system runs the traction or braking force, the Fr system, the Μ series car weight, and the α system acceleration (including the negative acceleration). Train driving resistance F r series When driving, for the convenience of calculation, usually only the following resistance is considered. Starting resistance: Resistance air resistance at the time of departure: The air resistance during train driving can simplify the system and can be removed or simplified. It can be reduced. Everyone is worried that because of the running support thrust, it will not limit the speed with the limit of the train. The main body, the dependent column, for each train needs a huge automatic train of time, sometimes using the following simple physical train driving resistance, that is, the resistance generated by the deceleration,

•10- 1284605 (7) • Slope resistance: slope of the route resistance curve resistance: curve resistance of the route 鹌 tunnel resistance: resistance air resistance generated when driving inside the tunnel, considering the resistance between the wheel tread and the track surface, usually Will adopt the speed of 2 times. In general, the train running resistance Fr is usually considered for the resistance φ force composed of slope resistance, air resistance, curve resistance, tunnel resistance, and starting resistance. Here, it is considered when driving a train other than a tunnel, so only slope resistance, air resistance, and curve resistance are considered. At this time, the slope resistance, the air resistance, and the curve resistance can be obtained by the following equations (8), (9), and (10) (for example, refer to the document "Operation Theory (DC AC Power Vehicle)" ). (a) slope resistance

Frg = s ... ( 8 )

Frg: slope resistance [kg weight / ton]

S: slope [%〇] (positive on the upslope and negative on the downhill) (b) air resistance

Fra = A + Bv + Cv2 ... ( 9 )

Fra: air resistance [kg weight / ton] A, B, C: coefficient v: speed [km / h] (c) curve resistance

Frc- 800/r ... ( 1 0 ) F rc: curve resistance [kg weight / 1 ο η ] -11 - (8) 1284605 . r · · radius of curvature [m] _ automatic train operation if using (7 In the model shown, even if it is an automatic train operation mode based on the driving plan, characteristics such as train characteristics and route characteristics will greatly affect ride comfort and stopping accuracy. [Summary of the Invention] [Means for Solving the Problem]

The invention is based on the premise that the train stops at a specific time at a specific time when the train is traveling between the stations, and the purpose thereof is to provide an automatic train running device and a train running support device, which can reduce the energy loss caused by driving and realize energy saving. Running. Moreover, the object of the present invention is to provide an automatic train running device, which can reduce the time and labor required for adjustment, and can automatically carry out the learning of characteristics after the driving, thereby further improving the ride comfort and increasing the stop. Precision. Further, it is an object of the present invention to provide a device for performing a necessary data collecting operation for the operation of a running device only when the train travels to and from a specific route. Further, an object of the present invention is to provide an automatic train running device which can realize: first, to eliminate the influence of chasing during automatic operation of the train, to improve the effect of saving energy; and second, to obtain the delay time, Improve the stop accuracy of the target position; third, it can improve the poor ride comfort caused by the change of the speed control command during the execution level operation. Moreover, the object of the present invention is to provide a train stop position automatic

-12 - 1284605 (9) The control unit can ensure stop accuracy _ without frequent switching of the level, and does not require a long adjustment period. In order to achieve the above object, an automatic train running device of the present invention is characterized in that it has a data processing means for processing the train driving data obtained online; and the train driving data obtained by using the data processing means and the previously acquired data are provided on the train. Automatically learn the control parameters of the train when driving, and the automatic characteristic learning means of the train characteristics and route characteristics; automatically execute the train automatic operation by using the train characteristics and route characteristics learned by using the automatic characteristic learning means Means of operation. Further, the automatic train running device of the present invention is characterized by having: a train characteristic learning means for collecting train characteristics and route characteristic information in train driving; and calculating a train optimum based on train related information collected by the aforementioned train characteristic learning means The operation mode, and the automatic train operation means for performing the automatic operation of the train according to this mode. φ [Embodiment] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Fig. 1 is a block diagram showing a schematic configuration of an automatic train operating device according to a first embodiment. Therefore, the embodiment of the system and the optimal train of the automatic train running device are particularly relevant to the "Lv Shi painting department", and the other parts are omitted. The optimal driving plan unit i 3 shown in Fig. 1 is composed of a driving mode compensation index calculating unit 15, a driving mode compensation unit 9, a driving distance compensation unit 20, and a timing determining unit 21. The driving mode compensation index calculation unit i 5 ' is composed of a loss index calculation unit 16 , an overload indicator calculation unit 丨 7 , and an addition - 13 · (10) 1284605 unit 18. The loss index calculation unit 16 calculates the loss index CPL ( X ) of the train position X based on the tentative driving mode (F0 stomach (X ), V0 ( X )). At this time, the CPL is Cost of Power Loss. At this time, the driving mode is expressed by the thrust Fn(x) and the speed Vn(x) of a certain position X.

Figures 2 and 3 are examples of various loss indicators. Figure 2 is the loss indicator for the operation, and Figure 3 is the loss indicator for the vehicle deceleration. In more detail, Fig. 2(a) is a machine loss indicator, Fig. 2(b) is a total loss indicator, Fig. 3(a) is a machine loss indicator, and Fig. 3(b) is a brake loss indicator. Figure 3 (c) is the total loss indicator. Here, the machine loss indicator refers to the loss index of the electric machine. Specifically, it is the addition indicator (frequency conversion transformer) loss indicator and the motor (main motor) loss indicator. These indicators are expressed as a function of velocity v and thrust F and are calculated by multiplying the loss [W] of a certain operating point (v, F) by the reciprocal of velocity [m/s]. By multiplying the reciprocal of the speed, a formal evaluation can be performed on the loss caused by the small velocity φ Δ v [m/s] at the velocity vl [m/s] of an operating point. The total loss indicator CPL (X) is calculated by multiplying the weight loss factor W1 by the total of the machine loss indicator and the braking loss indicator. The weighting factor W 1 is set from the viewpoint of the degree of loss reduction effect that can be obtained, or is set to be balanced with other indexes. That is, the loss index CPL (X) II Wlx (machine loss index + brake loss indicator) ... (2) The overload indicator calculation unit 17 is based on the tentative driving mode (F0 (X), V0 (11) 1284605 • ( x )) Calculate the overload indicator COL ( X ) of the train position x. COL is the Cost of Over Load 〇 Machine loss is the sum of converter losses and motor losses. Fig. 4 (a) and (b) are examples of converter loss [W] and motor loss [W] at each operating point. According to the tentative driving mode (F 0 (X), V 0 ( X )), the corresponding converter loss [W] and motor loss [W] are integrated respectively, and the converter loss plus the time between stations can be calculated. [J] and motor loss [J]. If φ is an overload exceeding the specification 値 [W], the corresponding overload indicator is calculated. For example, if the weighting factor is W2, the converter loss indicator COLC (X) can be COLC (X) = W2x {converter loss [J] / (travel time + stop time) a converter specification [W]} x conversion The loss indicator (Fig. 5(a)) is calculated by φ. COLC system Cost of Over Load in Converter 〇 Similarly, the motor loss indicator COLM ( X ) can be obtained using the motor loss indicator shown in Figure 5 (b) (however, the weighting factor is W3, which can be set separately). The overload indicator COL ( X ) can be added to these indicators, and is obtained by COL(x) = COLC(x) + COLM(x) ... ( 4 ) (8 -15- (12) 1284605. COLM system Cost Of Over Loss in Motor. Adder 1 8 adds the loss indicator CPL ( x ) and the overload indicator COL(x), and C(x) = CPL ( x ) + COL ( x ) to find the total train position x The index C ( x ). The driving mode compensation unit 19 adds the total index C ( X ) to the thrust mode F0 ( X ) of the tentative driving mode, and outputs the first compensation driving mode F0 1 (X ) (at this stage, The speed mode VO ( X ) does not change.) The first compensation driving mode F01 ( X) is only the compensator of the thrust mode, so the driving distance does not match the specific 値. In order to make the driving distance X and the specific 値 match, the driving The distance compensation unit 20 performs the compensation of the first compensation driving mode (F01 (X), V0 (X)) according to the route condition stored in the database 3, the vehicle condition, and the driving resistance, and outputs the second compensation driving mode (F02). (X), V02 (X)) and travel time T run. Distance compensation can be achieved by, for example, adjusting the taxi time. However, distance compensation The timing determination unit 21 determines whether the travel time T run is within the allowable error for a specific time. When the travel time T run is outside the allowable error, the second compensated travel mode (F02 (X)) , V02 (X)) is regarded as the new tentative driving mode (F0'(X), V0'(X)), and the calculation is re-executed. When the driving time T run is within the tolerance, it is regarded as the best driving 1246605 (13) • Mode (FI (x), V1 (x)) output. • Using the above configuration, tentative driving mode (F0 ( χ ), V0 ( χ )

The loss index CPL ( x) and the overload indicator COL ( x) can be used to compensate the thrust of the position x significantly. For example, Fig. 6 is a result of the driving mode of the embodiment of the present invention. At this time, it is assumed that the overload condition has not been reached, so the overload indicator has no effect. The "original mode (A)" is a tentative mode. The "Applicable indicator (B)" is the first compensation driving mode (F01(x), V01(x)). The higher the loss index is, the more the thrust compensation is implemented, and the weaker the braking force will be. On the other hand, although the running acceleration side is small, it also compensates for the traction force corresponding to the loss indicator. Although the "level quantization (C)" does not appear in the embodiment of the present invention, the level is only 6 segments, and if the continuous thrust cannot be output, the thrust F0 1 ( χ ) of the first compensation driving mode is selected. The thrust with the lowest thrust error. "Distance adjustment (D)" is the second compensation driving mode (F02 ( χ), V02 ( χ) that compensates the driving distance to a specific 値1 300 00 for the "level quantization (C)" mode. ). Compared with the loss of 2070 [kJ] in the driving mode before compensation, the loss of the second compensation driving mode is 1 65 0 [kJ], which reduces a considerable amount of energy loss. In terms of driving time, 84.5 [Sec] is compared with the former, and the latter is increased by 84. 9 [sec]. By repeating the calculation until the travel time becomes a specific time, it is possible to generate an optimum driving mode F 1 ( χ ) that minimizes energy loss in driving brake control while ensuring the timing and the fixed position stop. In this way, the best energy saving effect can be achieved while ensuring timing and positional stoppage. -17· (14) 1284605 - When only the driving mode in which the total energy loss is minimized is pursued, the variable voltage inverter 4 (converter) and the main motor 5 (motor) included in the driving brake device 2 may be used. The energy loss caused by electric machines increases. The operating range of the power machine is limited by the specifications. When the operating condition exceeds the specification, that is, when the overload condition occurs, the temperature rises due to heat generation, and the protective action or malfunction or burnout occurs. The overload indicator calculation unit 17 determines the degree of overload of each machine for the tentative driving mode. When the judgment result is overloaded, φ ’ will suppress the energy loss of the electric machine, and the thrust will be compensated for the overload indicator. Since the thrust compensation is performed from a region where the energy loss is large, the overload state can be effectively avoided. In this way, it is possible to avoid the stoppage and malfunction of the electric machine due to overloading, and improve the reliability of the system. Since the optimal driving plan is also implemented in the train, the position and speed of each moment can be used as the initial condition. In the case of ensuring the timing and stop position of the stop to the next station, the optimal energy saving driving mode φ is generated. That is, when the current driving mode is deviated due to the speed limit of the ATC or the like, the optimal energy saving driving mode can also be obtained from this state. If you follow the original driving mode, you may increase the loss and not the energy loss. Therefore, even in the event of an accident that deviates from the original driving mode, optimal energy saving can be achieved from that point in time. In the present embodiment, the position and speed are used as the initial conditions, and when the timing and the stop position are ensured until the next station, the optimal energy saving driving mode is generated, so that it can be applied not only to the automatic train between the stations. The automatic train running device (ΑΤΟ) that is operated can also be applied to the automatic train stop control device (TASC) that performs fixed-position parking control only in the area of the brake (15) 1284605. In addition, this embodiment is based on the premise that the driving distance and the specific 値 are matched, and the configuration is such that the calculation of the compensation of the driving mode is performed until the traveling time reaches a certain 値, and the configuration is also reversed. The calculation of the compensation for the driving mode is implemented on the premise that the driving distance is up to the specified level.

Figure 7 is a block diagram showing a schematic configuration of an automatic train running device according to a second embodiment, and the same portions as those in the first embodiment are denoted by the same reference numerals, and the description thereof will be omitted. Instructions are given. The database 3 inputs the operation schedule to the loss index calculation unit 16, and the database 36 inputs the operation load amount to the loss index calculation unit 16. The running load stored in the database 36 is the running load of the train in the acceleration of each feeding section at a certain time. The loss index calculation unit 16 extracts the corresponding operational load from the database information of the operation schedule and the operation load. As mentioned above, since the loss of the brake will change due to the running load, the loss index corresponding to the running load is calculated. Others are the same as in Figure 1. From the above, the following effects and effects can be obtained. Corresponding to the predicted operational load, adjust the loss indicator CPL ( X ), especially the brake loss indicator. For example, Fig. 3(b) shows the brake loss index when the load is fully operated. Because of the limitation of the capacitance of the inverter transformer 4, the higher the speed of the vehicle, the higher the loss index will be. Fig. 8 shows the vehicle loss index when the load is fully operated (125 kW / main motor). -19- (16) 1284605 • At this time, because the running load is insufficient, it is impossible to output the thrust command F cmd, which is equal to the area of the electric power. That is, the loss indicator will start to increase from the lower speed. Therefore, the energy loss caused by the load state can be reliably predicted, and more efficient energy saving driving can be realized. Fig. 9 is a block diagram showing a schematic configuration of an automatic train running device according to a third embodiment, and the same portions as those in Fig. 16 are denoted by the same reference numerals, and the description thereof will be omitted. Description.

The apparatus of Fig. 9 is provided with a database 34 and a driving mode separating unit 35 for replacing the tentative driving plan unit 1 2 and the optimal driving plan unit 13 of Fig. 48. The data bank 34 stores the driving mode at the time of driving between the stations of each train. The driving mode precipitation unit 35 extracts the driving mode F 1 ( X ) corresponding to the current inter-station driving from the database 3 storing the running time table. The driving mode stored in the database 34 can be realized by the following method, that is, the optimal driving pattern shown in the first embodiment is stored in advance, and the optimal driving mode of the result is stored. According to the above configuration, the following effects and effects can be obtained. In the generation of the optimal driving mode, it takes some time to perform the convergence calculation to perform the best I-picture. Therefore, when the driving of the next station is carried out during the parking at the departure station, the calculation time is sometimes limited and the optimum is not sufficient. Implementing these plans in advance avoids the limitation of computing time and gets the best driving mode. This way, you can further improve the energy saving effect. In addition, the driving mode is calculated in advance, and the driving mode can be accurately confirmed. In this way, the abnormal mode can be eliminated and the reliability of the system can be improved. -20- 1284605 (17) The block diagram of the electric train system of the train operation support device according to the fourth embodiment, and the same reference numerals as in Fig. 47 are attached to the same reference numerals and are omitted. Note, only the different parts are explained here. Here, there is provided a train operation support device 22 in place of the automatic train operation device 1 of the first embodiment. The train operation support device 22 is implemented and

In the same process as the automatic train running device 1 of the first embodiment, the thrust recommendation 値 Free is generated and output. That is, the train operation support device 22 outputs a thrust recommendation 値 Free for replacing the thrust command F cmd of the automatic train running device 1. This thrust recommendation 値Free is input to the thrust indicating device 24 provided to the main controller 23. The main controller 23 outputs a thrust command F cmd corresponding to the angle or position of the main controller to the drive brake device 2. The configuration of the thrust indicating device 24 is as shown in Fig. 11, for example. The thrust indicating device 24 is composed of an angle command calculating unit 25, an impedance controller 26, a servo amplifier 27, a servo motor 28, and an encoder 29. The servo motor 28 and the main controller 23 are mechanically coupled. The thrust recommendation 値Free outputted by the train operation support device 22 is input to the angle command computing unit 25. The angle command computing unit 25 calculates the angle of the main controller corresponding to the thrust recommendation 値Free of the input, and outputs it as an angle command Θ cmd. The impedance controller 26 inputs an angle command Θ cmd and the actual main controller angle θ detected by the encoder 29, and outputs the servo amplifier 27 such that the latter (angle Θ) and the former (angle command Θ cmd ) are The torque command T cmd of the purpose. The servo amplifier 27 drives the servo motor 28 in such a manner that the output torque of the servo motor 28 and the torque command T cmd are in the same manner. 〇-21 - (18) 1284605 - The impedance controller 26 is applied to the main controller 23 for the driver. The torque body pe controls the servo motor 28 in such a manner as to form a desired impedance (inertia moment J, damping D, stiffness K), and the block diagram of the control system is as shown in FIG. The total coherence moment of the rotor of the J 伺服 servomotor 28 and the main controller 23, gl and g2 are equivalent to the cutoff frequency of the filter for the purpose of removing interference. When the angle command Θ cmd is zero, the torque applied from the outside to the main controller 23, that is, the torque applied by the driver to the main controller 23 reaches the main φ controller angle 传达 the transfer function 0(s) If the interference cutoff filter is ignored, it can be expressed as follows, so it is known that the desired impedance (J, D, K) can be obtained.

1 J^s2+D^s-l·K

Tope

The above composition has the following effects and effects. The thrust indicating device 24 controls the angle φ degree 主 of the main controller 23 with the servo motor 28 to obtain a thrust command F cmd which is calculated by the train operation support device 22 and which is free. In this way, when the driver operates the main controller 23, it is controlled by the impedance of the impedance controller 26, so that the driver feels that the desired impedance (J, D, K) has been reached. That is, the driver can obtain the thrust command F cmd corresponding to the thrust recommendation 値Free without the main controller 23 being touched. On the other hand, when the driver operates the main controller 23, it can withstand the force from the servo motor 28 to the thrust in the direction of the thrust, and can be set at any angle, that is, the thrust command F cmd. That is, the driver can also entrust the driving to the train operation support device 22, and in the case of the necessary -22-(19) 1284605, the driver operates the main controller 23 and controls the thrust command according to the consciousness to save energy. The angle Θ of the main controller 23 for the purpose of operation can be detected by the reaction force from the main controller 23, and the driving can be performed with the realization of the energy saving position stop mode. Therefore, in addition to the use of the driver's operation to achieve energy-saving driving and positioning to stop driving, in the event of an unexpected situation, you can quickly take countermeasures.

The thrust command F c m d for driving the brake device 2 is not directly controlled by the train operation support device 22, but is provided by the angle Θ of the existing main controller 23, and the system can be simplified. Further, in the case of the train operation support device, it is necessary to pass the driver afterwards, and it is not necessary to require the train operation support device 22 to have strict positional stop accuracy, so that the device can be simplified. In this way, the reliability and cost of the system can be improved. Moreover, the train operation support device needs the driver after all, so the driver's operation technique needs to be required at any time. According to this embodiment, it is possible to avoid the problem that, in the case of a system having an automatic train running device, it is possible to cope with an unexpected problem due to a decrease in the operating technique of the driver. Fig. 3 is a block diagram showing a schematic configuration of a train operation support device according to a fifth embodiment. Since the configuration of the thrust indicating device 24 is different from that of the fourth embodiment in the present embodiment, the different portions will be described here. However, in the present embodiment, the thrust command is accelerated in six stages (P 1 to P6 ), the brake is decelerated in six stages (B1 to B6), and the neutral position (N), and the emergency brake (EB) is adopted as the main controller level. the way. The grade here refers to the model used to control the speed and thrust, and is used in the current tram drive control. The number of segments can range from several segments to 30 segments -23- (20) 1284605 and above, depending on the system. Further, the main control unit 23 of Fig. 3 is a schematic configuration when viewed from above.

The thrust indicating device 24 is composed of a recommended level indicating the control unit 30 and the lamp group 31. In the illustrated embodiment, the lamp group 31 is composed of six lamps corresponding to the operation acceleration levels P1 to P6, six lamps corresponding to the brake deceleration levels B1 to B6, lamps corresponding to the neutral level N, and corresponding emergency brakes. The EB lamp is composed of 14 lamps. The recommended level indicates that the control unit 30 receives the recommended level command N rec of the train operation support device 22, and executes the control in which the corresponding lamp is lit. According to the above configuration, the following effects and effects can be obtained. The driver can use the lighting to confirm whether or not the level is set to achieve the goal of saving energy in the case of ensuring the timing and the stop position. For example, the content of the recommended level command N rec is the running acceleration level P6, and its corresponding lamp will illuminate, and for the braking deceleration level B3, the corresponding lamp will illuminate. The driver observes the condition of the lighting, performs the level operation of the corresponding main controller 23, and realizes the energy saving operation for suppressing the energy loss. There is no direct electrical/mechanical connection between the thrust indicating device 24 and the driving brake control system, and the driver's operation is required. Therefore, in the event of an unexpected situation, the driver can quickly respond according to the judgment of the driver, and the system is improved. Trustworthiness. The lamp and the display device using the LED (Light Emitting Diode) are easier to implement than the servo mechanism of the main controller 23 of the fourth embodiment, and the reliability of the system can be improved, and the cost of the device can be further reduced. Fig. 14 is a schematic diagram of a train operation support device according to a sixth embodiment - 24 - (21) 1284605 - a block diagram of an example. The present embodiment differs from the fifth embodiment only in the configuration of the thrust indicating device 24. Therefore, only the different portions will be described here. The thrust indicating device 24 of the present embodiment indicates the control portion 32 by the recommended level. And the sound output unit 33 is configured. When the recommended level command N rec is received from the train operation support device 22, the recommended level indicating control unit 3 2 controls the sound output unit 33 to output the corresponding voice. For example, if the recommended level φ is B3, a voice such as "Block 3" will be issued. With the above configuration, the following effects and effects can be obtained. The driver can know by voice the level of energy saving driving in the case of ensuring timing and positional cessation. According to this aspect, the same effects and effects as those of the fifth embodiment can be achieved. When the recommended level is indicated by a lamp in the fifth embodiment, the driver's attention will be focused on the result of the indication, or an accident may occur due to failure to pay attention to the front. In contrast to the use of voice instructions, you can avoid this problem and improve the system's letterworthiness. Fig. 15 and Fig. 16 show an embodiment of the automatic train running device of the present invention. The automatic train running device (ΑΤΟ) 1〇8, which is placed on the train 0 shown in the figure, obtains the speed limit data from the automatic train control device (ATC) 102 of the above ground system, and also from the database in the train 0 (D Β ) 1 0 3 Obtaining route conditions (inclination angle and radius of curvature of the curve, etc.), vehicle conditions (number of trains, weight, etc.), and operating conditions, etc., and also obtaining a departure signal from the driver's station 104, from the load-carrying device 105 The load signal is obtained, the train speed signal is obtained from the speed detector 106, and the train/stand signal is obtained from the above-ground sub-detector 107 that responds to the ground level of -25- 1 284 605 (22). . The above-ground sub-systems that are appropriately placed on the route are used to confirm the train position. Here, the DB 103 indicates that it is placed in the train 0, and may be a ground system located outside the train, or may be distributed in the train and on the ground.

In addition to the data processing means 180 for performing on-line data processing and the automatic train operation means 181, the ΑΤ0 108 includes estimation means and learning means represented by the pre-employment characteristic estimating means 124 and the post-business characteristic learning means 134 which will be described later. The data processing means 180 will process the train speed signal, in addition to the processing of the train speed, the train position (time integral of the speed 値), the train acceleration (the differential 速度 of the speed), and the train travel distance (the time of the absolute speed) The integral 値) implements continuous operations. From the train position to the train travel distance, moderate compensation is performed according to the train position signal of the ground sub-detector 107. The data processing means 180 performs a specific calculation based on each input signal, and provides the measurement data necessary for the learning and train self-operation described later. The necessary measurement data for the automatic operation of the train will be provided to the train automatic operation means 81. The train automatic operation means 1 8 1 outputs a running command to the drive unit 9 or outputs a deceleration command to the speed reducer 1 1 based on the result of calculation using each input data. The drive unit 109 includes a main motor for the purpose of towing the train, and a power converter for controlling the same. Further, the reduction gear unit 110 usually has both a mechanical brake and an electric brake. ΑΤ0 108 is placed on the train 0, and the part of the pre-business characteristic estimation means 124 and the post-business characteristic learning means 134 related to the learning of the present invention is detailed in the figure -26- 1284605 (23) • 16 The pre-business driving judgment means 〇2, the business pre-characteristic initial setting means 1 21, the pre-business test driving automatic train operation means 122, the driving result storage means 123, and the pre-service characteristic estimation means 1 2 4 Estimation means compensation means 1 2 5, characteristic estimation 値 storage means 1 2 6 , learning characteristic database (learning characteristic DB) 1 3 0, characteristic initial setting means 1 3 1, automatic train operation means 1 32, driving after business The result storage means 1 3 3, the post-business characteristic learning means 1 3 4, and the learning result compensation means φ 135 are comprised. The means 121 to 126 are processing means for the purpose of driving the vehicle before the driving test, and the means 1 3 1 to 1 3 5 are the processing means for the purpose of driving, the pre-business driving determination means 120 and the learning characteristic DB 130 are It is irrelevant before and after the business trip, but it is set up in a way that is shared by both. In Fig. 16, the data processing means 180 and the automatic train operation means 181 which are originally used as the automatic train running device are omitted. Next, the operation of the devices of Figs. 15 and 16 will be described.

In Fig. 15, ΑΤ0108 will obtain the speed-restricted data from ATCl〇2 in advance, obtain the pre-acquired information such as route conditions, vehicle conditions, and operating conditions from DB1 03, and simultaneously obtain the speed, and then perform a specific operation to generate The control command consisting of the running command or the deceleration command realizes the automatic operation of the train 如 as described above. The AT0108 receives the departure signal from the bridge 104 and starts the automatic operation using the automatic train operation means. After the departure, the load information obtained from the load-carrying device 105, the speed data obtained from the speed detector 1〇6, and the ground sub-test obtained from the above-ground sub-detector 1 〇 7 are used. 27-(24) 1284605 Ii [J information. The load information is used as the weight related information of the train, and the above ground detection information is used for the compensation of the position information. Using these information' ΑΤΟ 108, the train's control command (running command/deceleration command) can be drawn. ° When the running command is prepared as a control command, the running command will be output and utilized.

The drive unit 1 0 9 causes the train to operate. In addition to the running torque (running traction) command, the running command has a running level command and the like when driving. Further, when the deceleration command is prepared as a control command, the "deceleration command is output" is used to decelerate the train by the deceleration device 1 10 . The deceleration command is the braking force command' level. When driving, it is the braking level command. Next, a detailed description of the action of ΑΤ0 10 8 will be described with reference to Fig. 16. When receiving the departure signal from the driver's station 104, first, the pre-business test driving means or the driving after the business is judged by the pre-employment driving determination means 120. The method of judging at this time may be a method using a flexible flag such as "testing when the flag is not set", "having a business trip when the flag is established", and a method of setting the result by using a rigid switch. When the pre-business driving determination means 120 determines that the pre-business test is being carried out, the pre-business characteristic initial setting means 1 2 1 sets the initial characteristic parameters at the time of the pre-business test driving. The method of setting can be considered by using the human-machine interface to manually implement the setting method before the start of driving. In addition, it is only necessary to input the characteristic parameters from the information that can be obtained in advance, such as the specifications and route characteristics of the train. Then, the test vehicle using the automatic operation train is implemented by the pre-service test train automatic operation means i 22 using the characteristic parameter set by the pre-service characteristic initial setting means 1 2 1 '. The automatic train operation side rs -28- (25) 1284605. In terms of law, if the best driving plan is drawn when parking at the station, according to the actual @, automatic operation, and the best driving plan has a large deviation, re A plan for planning a driving plan or a compensation for the use of error feedback for a control command. Here, a test for the use level for the purpose of characteristic estimation is performed for a pre-operating pre-operating vehicle, for example, a train traveling at a level. Driving, etc., the stomach performs the purpose of the calculation of the characteristics.

Next, the result of the automatic operation of the train automatic operation means 1 22 #1 before the pre-business test is stored by the driving result storage means 1 23 . When storing, the target driving plan, speed information and location data measured at the time of driving are regarded as electronic files stored in a medium such as a hard disk (HD). Next, the calculation of the characteristic parameters is carried out by the pre-service characteristic estimating means 1 24 using the test driving result stored by the driving result storage means 123. Predicted characteristic parameters such as weight, acceleration characteristics, and deceleration characteristics should be implemented before the business. In terms of the weight of the number of trains, the number of trains is based on the test drive before the operation. Therefore, no passengers can ride, and the acceleration or deceleration during taxiing and the driving resistance of the train can be used to estimate. Here, a case where the train of the object is expressed by the simple physical form of the equation (7) is considered. In terms of train running resistance, calculations can be performed using the formulas of the route characteristics such as slope and curvature, air resistance, and frictional resistance. In addition, please refer to the document "Operation Theory (DC AC Power Station)" for the calculation of train running resistance. In general, the train running resistance Fr can be expressed by the following formula. (s) -29- 1284605 (26)

Fr = Fr g + Fra + Frc = s + (A+Bv+Cv2 ) + 800/r ".(11)

However, Fr is the train resistance [kg weight / ton], Frg is the slope resistance [kg weight / ton] (upward slope is positive, downhill is negative), Fra is driving resistance [kg weight / ton], Frc is curve resistance [kg weight / ton], s is the slope [%. ], A, B, and C are coefficients, v is the train speed, and r is the radius of curvature. If the above items are considered, the weight can be estimated by the variant of equation (7). M = ( F - Fr ) la ... ( 12) In the formula (12), when the vehicle is coasting, the running traction force F may be 〇 (zero). Further, in terms of acceleration (or deceleration) α, calculation can be performed using the measurement result (train travel speed) by the most φ flat method or the like. In the above process, the weight Μ can be derived. After the calculation of the weight Μ is completed, the weight estimation 値 can be used to estimate the running characteristics and the braking characteristics. First, use the weight estimation 値Mest, the acceleration a acc at the running time, and the train running resistance Fr to estimate the operating characteristics (the relationship between the operating level and the running traction, etc.). The acceleration a acc and the train resistance Fr in operation can be obtained by the same processing as the aforementioned weight calculation. Using the weight and weight calculation, the running traction force F can be estimated by the following formula.

Cs) -30- 1284605 (27)

13 F = Mest a acc + Fr When the train is operated with the grade, the running traction of each grade can be calculated by equation (13). It can also be used to estimate the relationship between the operating level and the running traction.

In addition, the braking force characteristics can be estimated by using the weight estimation 减, the deceleration during deceleration, and the train running resistance. The deceleration at the time of deceleration and the train resistance can be obtained by the same processing as the above weight calculation. Using this and the weight calculation 値, the braking force F can be estimated by the following formula. ...(14) F = Mesta dec + Fr However, adec is deceleration (negative acceleration). When a train that performs a brake operation is used, it is possible to calculate the vehicle power of each class of φ by the formula (1 4 ). This result can be used to estimate the relationship between the brake class and the braking force. These calculations are best calculated after the station is driving or when the vehicle is parked. However, it is also possible to perform calculations in the train and confirm the calculation results in the train. By implementing the weight, the running characteristics, and the braking characteristics in this way, the number of trains can be adjusted in the number of trains, and the adjustment can be completed in a shorter period of time before the business trip. Next, the characteristic obtained by the pre-business characteristic estimating means 124 is estimated, and the compensation result compensation means 125 is used for compensation. When performing compensation (28) 1284605 λ, it should be set within the allowable range of theoretically achievable characteristic parameters and must be corrected to this allowable range. For example, if the characteristic estimation is beyond the allowable range, it is conceivable to use the setting of the calculation beforehand, or the restriction within the allowable range. If the deviation from this allowable range is too large, the operation of the test driving or the like must be re-executed. Next, the characteristic calculation of the compensation by the estimation result compensation means 156 is performed, and the characteristic estimation calculation means 126 is stored in the learning characteristic DB 130. The method of storing can be the same as the above-described driving result storage means 1 23 . The learning characteristic DB 130 can store the characteristic learning results obtained after the business trip described later, in addition to the characteristic calculation results obtained by the trial driving before the driving. The case where it is determined by the pre-vehicle driving determination means 120 that it is driving after the business is described below. When driving, the initial parameters of the characteristic parameters are set first by the characteristic initial setting means 1 31. At the time of the initial business trip, the characteristic parameters (characteristic estimation results) stored in the storage means 126 are estimated using the utilization characteristics obtained from the learning characteristic DB13 0. When learning is carried out at the same time as the business travels, the characteristic parameters obtained from the learning results (characteristic learning results) can be used. Next, using the characteristic parameters set by the characteristic initial setting means 1 31, the train automatic operation means 1 32 Perform automatic running of the train. In terms of the automatic operation of the train, basically, it is the same as the automatic train operation means 122 for the pre-operating test train. When there is an unspecified number of passengers, the weight changes. Therefore, the initial shipment from the station -32- 1284605 (29). When you travel, you must estimate the weight of the station when driving. For the method of weight calculation, if _ can obtain the load, the load should also be used. When the load is not available, _ can be used to calculate the weight in the same manner as the pre-service characteristic estimation-means 124 and the estimation result compensation means 125 at the initial stage of the station departure. If the result of the calculation and the characteristic initial setting means 1 3 1 are different, the processing plan and the like must be re-executed. Figure 17 is a summary of the weight calculations performed during the initial operation from the station.

In Fig. 17, the horizontal axis is the distance from the departure station to the next station, that is, the position. The vertical axis indicates the speed of each position in the speed mode. According to the characteristics of the departure station, the optimal driving mode is calculated according to the characteristics of the departure. 1 3 1 (fine dotted line) After the execution of the driving, the actual driving result according to the initial running interval 1 3 0 is the actual driving mode i 3 2 (thick The solid line) implements the weight calculation, and based on the weight calculation, calculates the driving mode 1 3 2 (thick dotted line) by recalculating and implementing the compensation, and implements the actual running operation accordingly. Next, the result of the automatic operation performed by the automatic train operation means 32 is stored by the post-business driving result storage means 33. The method of storing ' can be the same as the above-described driving result storage means 23. Next, the characteristic learning is performed by the post-business characteristic learning means 134 using the driving result stored in the post-business driving result storage means 133. Regular learning aspects of this feature are implemented in the following manner. (1) Learning based on the results of the inter-station driving (2) Learning based on the results of the whole route (3) Learning based on the results of the 1-day driving (4) Learning based on the results of several days of driving

iR -33- (30) 1284605 • (5) Learning based on several months of driving results. The following is a description of each of (1) to (5) above. • (1) Learning based on the results of the inter-station driving According to the inter-station driving results obtained after the inter-station driving, the learning results are reflected in the next station driving. For example, when it starts to rain, learn the correspondence when the braking force is reduced. It is judged that it is necessary to perform an example of learning for the driving result φ of one station, for example, the corresponding reduction in the vehicle power in the rainy day. In rainy days, if the train uses air brakes, the rain will reduce the friction of the brake blocks and reduce the braking force (deceleration performance). At this point, after the start of rain, you should find that the deceleration performance is reduced. Just learn the characteristics of the braking force based on this result. The learning result at this time is usually temporary, so it can be stored separately and used as a temporary characteristic parameter. (2) Learning based on the results of the whole route

Perform the learning based on the driving results from the first to the last of the route, and reflect the learning results on the driving of the next route. For example, when there is almost no shortage of the target stop position (deviation amount) at the end of a route, in order to eliminate the amount of deviation, it is sufficient to perform the learning of the vehicle force characteristic corresponding to the deviation amount. For example, when the target stop position is exceeded, the setting of the braking force characteristic should be slightly larger than the actual 値. That is, the assumed deceleration cannot be obtained because it is larger than the actual braking force. In this case, it is only necessary to carry out the learning in which the setting of the braking force characteristic is slightly smaller. -34- 1284605 (31) • (3) Learning based on 1-day driving results. Perform learning based on 1 driving result and reflect the learning results on the next day's driving. For example, when observing the results of the driving on the 1st (for example, 1 - the driving result of several trips of the whole route), it can almost be said that the parking between the stations will definitely exceed the same as the target stop position. The degree is likely to be an error in the setting of the route characteristic parameters such as the slope and curve between the stations. In this case, it is sufficient to perform the learning of the route characteristic parameters such as the slope and the curve with a slight φ corresponding to the driving result. (4) Learning based on several-day driving results Store the results of several days of driving and perform learning based on the stored results. For example, 'observing the results of several days of driving, if the deviation of the driving plan will only occur at the same time, it should be affected by certain factors, and only the running traction characteristics or braking force characteristics of the time zone are deviated. The actual situation. If there is no deviation in other time zones, the characteristic parameter itself φ should not deviate from the actual, so only the characteristics of the object time zone are compensated, and then the compensation 修正 can be corrected by using the learning. (5) Learning based on the results of several months of driving # When several driving results in January are used, the learning is performed based on the stored results. For example, 'Learning based on the results of the storage, such as during the maintenance check. For example, by observing the results of driving in three months, it can be seen that the vehicle power of 3 months ago, 2 months ago, and 1 month ago will gradually decrease as time passes. In this situation, it is difficult to learn from the results of several days of driving.

-35· 1284605 (32) . To judge. When using an air brake, it is likely that friction causes the brake block to wear. Therefore, it is necessary to change (learn) the characteristic parameters based on the result, or to take measures such as replacing the brake block according to the degree. In addition, it is also possible to use measures to change the aging time such as the wheel diameter. In the above study, the examples shown in the flowchart of Figure 18 can be selectively used to implement the learning. In Fig. 18, the pre-business driving judgment means 120 is used to carry out the judgment of the pre-business test driving or the business driving after the business (steps)

1 5 1 ) When the result of the judgment is the former (pre-business test driving), the pre-employment test driving is carried out (step 152), the initial parameter estimation is performed (step 135), and the processing is terminated. If the result of the determination in step 151 is "business driving", one of the five kinds of learning corresponding to the driving content is implemented. That is to say, 'determine the form of driving at the end of the business trip (step 1 54), and if it is to end the inter-station driving, implement "(1) learning based on the results of the inter-station driving" (step 1 5 5), if it is to end the full-route driving Then, "(2) Learning based on the results of the entire route is implemented" (step 156). In step 154, if the vehicle is 0 for the end of the day, it will further determine how many times the data is stored (step 1 5 7). According to the result of the judgment, if the data for one day has been stored, "(1 day) "Learning of driving results" (Step 1 5 8), if it is a stored number of days of data, implement "(4) Learning based on several-day driving results" (Step 1 59), if it is stored for several months Then, "(5) learning based on the driving results of several months" (step 160) is implemented. However, each learning step indicated by thick lines in Fig. 18 is 1 5 5, 1 5 6 , 1 5 8 , 1 5 9 , 160, will only be implemented when the driving result shows the tendency to learn as shown below. That is, -36- © (33) 1284605 • a ) Continue to present deviations of the same tendency (for example, full route Driving • In the result, 'the same degree of target stop position is exceeded when all stations are crossed, etc.'; and ** b) when there is a significant deviation. Learning can be considered as a method of increasing or decreasing a certain characteristic parameter by a certain ratio. For example, as mentioned above, in the whole route driving result, when all the stations have the same degree of target stop position exceeding, the setting of the braking force φ should be slightly larger than the actual braking force, so the implementation of reducing the braking force by a certain ratio The setting of the characteristics of the learning. In particular, depending on the learning outcomes of the inter-station driving results, there are few cases where deviations of the same tendency occur. Therefore, at this time, the following learning should be implemented. In other words, • Automatic train operation mode: When there is a considerable deviation between the driving plan and the actual measurement, the automatic train operation mode for compensation of the control command (running level command, brake level command, etc.) is implemented corresponding to the difference in φ. • Learning method: When there is a deviation between the driving plan and the actual measurement, the learning is performed in response to the compensation of the control command. Taking the braking force characteristic as an example, for example, when braking, if there is a control command compensation that is greater than the plan, it should be a deceleration that is not assumed. In this case, the setting of the braking force characteristic should be too large, so it is sufficient to implement the setting of reducing the setting of the braking force characteristic by a certain ratio -37- (34) 1284605. If there is a control command that compensates for the brake level less than the plan, the opposite is to perform the learning of setting the brake force characteristic at a certain rate. The judgment of the difference between the estimated characteristics and the actual , is based on the acceleration and deceleration obtained in the form of measurement data, using the characteristics of the train, the characteristics of the route shape (slope, curve, etc.), the weight, the running traction or the braking force. To determine whether the formula (7) is satisfied. As described above, the learning result compensation means 135 performs compensation for the result of the learning by the post-business characteristic learning means 134. The method of compensation ' can be performed in the same manner as the above-described estimation result compensation means 1 25 . This compensation result is stored as a characteristic learning result and stored in the learning characteristic DB130. As shown above, learning is carried out even during business operations, and business operations are performed while adjusting the characteristic parameters. Most of the above studies are automatically learned on the line φ in the train stop at the time of the station. However, the calculation of the weight of the runtime is automatically calculated on the line in the vehicle. In this way, automatic operation can be carried out by using the continuous implementation of learning and calculation of the automatic operation of the train, which can be performed in a situation where the number of trains is different and the timeliness changes are well matched. As described above, with the automatic train running device of the seventh embodiment, the calculation of the weight, the running traction, and the braking force can be performed before the driving. For the number of trains of different trains, it can be adjusted in a shorter period of time than before. After the operation, the characteristic parameters can be learned, so even the characteristics are

-38- 1284605 (35) When there is a change, it can still achieve good ride comfort and stop accuracy. 趵 Automatic operation. In addition, after the business, we can learn more about the actual situation based on the use of the data during the period--------------------------------------------------------------------------------------------------------------------- In addition, in the calculation before the business and in the post-business study, compensation for the calculation and learning results will be implemented. In the event of an impossible result, it can be compensated without using the impossible characteristic parameters. Implement calculations and learning.

In the above way, as the characteristics learning progresses, an effective driving plan can be developed. In addition, if there is a large learning in the train, the learning will be triggered, and the driving plan will be redesigned to realize the automatic train operation that satisfies the ride comfort, the stopping accuracy of the target stop position, and the driving time. In the seventh embodiment, most of the learning is automatically learned on the line in the train stop when the train arrives at the station, and the weight calculation at the time of the train is automatically calculated on the line in the train. However, if there is a human-machine interface in which the learning progress can be confirmed in the train, automatic online learning can be performed in the driving, and the system using the learning result can be realized at the judgment of the driver. At this time, it is also possible to use only the learning means as a separate device and use it as a support device for automatic train operation. Fig. 19 is a view showing an essential part of the automatic train running device of the ninth embodiment. In this embodiment, the post-business characteristic learning means includes an automatic characteristic learning means 1341 for each request item, an automatic characteristic learning means 1342, an automatic characteristic learning means 1343, an automatic characteristic learning means 1344, and an automatic characteristic learning means 1345. In addition, 尙 has the input of these automatic characteristics learning means (S) -39 - 1284605 (36) • The learning result comparison means 1 3 6 of the obtained learning result, and the comparison result according to the learning and learning result comparison means 1 3 6 Compensation measures for learning outcomes that compensate for learning outcomes 1 3 7. - The automatic characteristic learning means 1341 to 1 3 45 perform the characteristic learning as described in the seventh embodiment. The learning result comparison means 136 accepts the learning results of the automatic characteristic learning means 1341 to 1345, compares the learning results, and checks whether there is a large contradiction between them. In the automatic characterization method 1 3 4 1~1 3 4 5, there is a considerable difference in the interval between the learning periods, and basically, the length of the learning period is longer than the one in the learning period. The result of one party can be. For example, the learning result of the automatic characteristic learning means 1 3 4 5 is obviously the same as the n-time of the learning result of the automatic characteristic learning means 1 3 44 of the same time zone, for example, 10 times, it is judged as a significant abnormality, and will be automatically The learning result of the characteristic learning means 1 345 can be regarded as the result of a major contradiction. Moreover, the inspection is performed by the complex result in the automatic characteristic learning means 1 3 4 1 to 1345, and the inspection accuracy can be further improved.

Secondly, the learning result compensation means 137 will perform compensation for the comparison result of the major contradiction in the learning result comparison means. In the method of compensation, the easiest method is to directly use the learning result of the automatic feature learning method with a short learning period (learning interval). However, when using the complex learning results of the automatic characteristic learning means 1341 to 1345, the average 値 of these learning results can also be considered. Moreover, if the learning results of most of the automatic characteristic learning means 1341 to 1345 appear to be contradictory, or the learning results of the automatic characteristic learning means 1341 to 1345 are large (37) 1284605. Consider using its average 値. • Automated feature learning means 1 3 4 The adaptive observer can be used to perform the feature • Learning. If the target device has been modeled as shown in equation (7), the observer should use the observable (detection) to identify the parameter. The system identification can also be carried out by type, and the train automatic operation means 81 can utilize the identification result of the adaptive observer at any time to form an adaptive control system. In the case of equation (7), the adaptive observer can be used to observe the acceleration and deceleration of the crucible (calculated from the detection speed of the speed φ detector 106), and the running traction or braking force of the control command, to identify the weight and the driving resistance at any time. . For the calculus of the observer, you can use the method of expanding the least squares, expanding the Kalman observer, and adapting to the observer. (For details, please refer to the "Introduction to Strong Adaptive Control" (Taiji Manju, Jinjing Ximeixiong, OHMSHA) Chapter 2 "Inferred and Adapted Observers for Unknown Devices" P.47~87, or "System Control Series 6 Best Filtering" (Xishan Jinglu, Peifeng) Section 3.3 "Adapting to the Viewer" Ρ·50~57 ).

As shown above, the comparison of several automatic characteristic learning means with different learning periods (learning intervals) is carried out to eliminate the contradictory learning results, and a more accurate characteristic learning result can be obtained. In the eleventh embodiment, the automatic characteristic learning means 134 can also perform characteristic learning using the interference observer. Interference observers use motion control, etc., to identify interferences (for details, please refer to "Designing with MATLAB Control System" (edited by Noda Kenji, Nishimura Hideo and Hirata Hikaru, Tokyo University of Electrical Engineering Publishing House) Section 4.4" Motion Control Interference Observer "Ρ.99~102). The train running resistance of the formula (1) is regarded as the motion control -41 - (38) 1284605 - The force interference can be calculated at any time by using the interference observer. - Using this calculation result to implement learning, you can perform learning with higher precision. A detailed description of the first embodiment of the present invention will be made with reference to the drawings. Fig. 20 is a configuration diagram of the automatic train running device 1 and the data storage unit 201. The automatic train running device 1 is composed of a train characteristic learning device 207 of a train characteristic learning means and an automatic operation control unit 208 of an automatic train operating means. The train characteristic learning device 207 obtains the characteristic data (train resistance, delay time, etc. (described later)) and route information of the train φ in the train. The data acquired by the train characteristic learning device 207 is stored in the data storage unit 201. The data acquired by the train characteristic learning device 207 and stored in the data storage unit 201 is output to the automatic operation control unit 208. The automatic operation control unit 20 8 formulates a driving plan based on the data acquired by the train characteristic learning device 207 and stored in the data storage unit 20 1 . The train will operate automatically according to this driving plan. The train characteristic learning device 20 7 is a train storage device φ 20 1 , a train weight calculation means, a train weight calculation unit 209 that operates the traction force deviation detection means, and a train resistance calculation unit 2 1 0 of the train resistance calculation means. The calculation means and the braking force calculation means 2 1 1 , the delay time calculating means 2 1 2, and the riding rate calculating means 2丨3 of the riding rate calculating means, and the detection of the train speed Composition. The output of the data storage unit 201 is input to the train weight calculating unit 209, the train resistance calculating unit 2 1 0, the braking force calculating unit 2 1 1 , the riding ratio calculating unit 213, and the automatic operation control unit 208. Output of train weight calculation unit 209

-42- (39) 1284605 • It is input to the data storage unit 20 1 . The output of the train resistance calculating unit 2 1 0 is recommended to the data storage unit 2 0 1 . The output of the braking force calculation unit 2 1 1 is input to the data storage unit 201. The output of the delay time calculation unit 212 is input to the data storage unit 201. The output of the ride rate calculation unit 213 is input to the data storage unit 201. The output of the operation control unit 8 is input to the train weight calculation unit 209, the braking force calculation unit 2 1 1 , the delay time calculation unit 2 1 2, and the ride rate calculation unit 2 1 3 .

When the train acceleration operation is performed, the data storage unit 211 inputs the train resistance 値 and the automatic operation control unit 208 to the train weight calculation unit 209 to operate the traction force 値F and the train speed V at the current point. The train weight calculation unit 209 calculates the train weight 以 using the train resistance 値Fr, the running traction force 値F, and the train speed V by Equation 15. The train weight 求 obtained by the train weight calculating unit 209 stores the data storage unit. In Equation 15, Μ is the train weight, F is the running traction 値, Fr is the train resistance 値, and α is the train acceleration. The train acceleration α can be obtained by using the train speed V. (15) M = (F - Fr) / a The train weight calculating unit 209 can also be used as the operating traction force deviation detecting means for the running traction force 値F, and the train weight 部 calculated by the train weight calculating unit 209 can be used as the speed V値, and when the 値V 1 used to calculate the train weight 不同 is different, it can be substituted into Equation 15 to obtain the correct running traction force 値F. The train weight calculating unit 209 can also detect the deviation between the running traction force 値F and the running traction force finger (40) 1284605 - 値Fk indicated by the automatic operation control unit 208. The deviation of the running traction command 値 “and the running traction force 値F is output to the data storage unit 20 1 for storage. Since the deviation of the traversable operation • traction command 値Fk and the running traction force 値F can be detected, the traction command can be operated during the detection.値Fk plus the deviation of the running traction command 値Fk and the running traction 値F, can be calculated as the new running traction command 値Fk 'this treatment' can achieve more accurate automatic train operation. When the train taxis, the data storage department 20 i inputs the train weight Μ and the speed V to the train resistance calculating unit 2 i 〇 φ. The train weight 値Fr can be calculated by the formula i 6 using the train weight Μ and the speed V input by the data storage unit 20 1 . The traction force is not running, so the running traction force 値F is 0. Since the running traction force 値F is 0, the formula 15 can be deformed to derive the formula 16. The train resistance 値Fr calculated by the formula 16 is output to the data storage unit and stored. In 1 6 , Μ is the train weight, F is the running traction 値, Fr is the train resistance 値, α is the train acceleration. The train acceleration α can use the train. V degrees strike.

Fr=F— Μα = 0- Μα (16) Train resistance 値Fr as shown in “Operation Theory (DC AC Power Vehicle), Dating Society, etc.), when the general train (there is a difference in high-speed vehicles), the slope resistance The sum of 値Frg, curve resistance 値Frc, and driving resistance 値Fra can be expressed by Equation 17. Further, it is understood that the slope resistance 値Frg, the running resistance 値Fra, and the curve resistance 値Frc can also be expressed by Equation 18, Formula 19, and Formula 20, respectively. -44- (41) 1284605 - Since the train resistance 値Fr during taxiing can be calculated by the train weight Μ and the speed • V, the train resistance calculating unit 210 can also calculate the slope resistance 値Frg and the driving resistance 値Fra. The driving resistance 値Fra can be calculated using the speed V. _ Again, the curve resistance 値Frc uses the data previously stored in the data storage unit 1. Since the train resistance 値Fr, the driving resistance 値Fr, and the curve resistance 値Frc can be used as the data, the train resistance calculating unit 210 can calculate the slope resistance 値Frg using the deformation of the formula 17. The obtained φ slope resistance 値Frg calculated by the train resistance 2 1 0 is output to the data storage unit 201 and stored. In Equation 18, the s-system slope [%] is positive for uphill and negative for downhill. In Equation 19, the coefficients of the A, B, and C systems and the velocity of the V system are [km/h]. In Equation 20, :r is the radius of the curve [m]. The train resistance calculation unit can detect the slope resistance and the train resistance 在 when the train is driving, so the correct data can be detected. Moreover, as long as the data is detected by performing a round trip on the planned route, it has a considerable time reduction effect. In Equation 17, Equation 18, Equation 19, and Equation 20, the train resistance FFr, the slope resistance FFrg, the driving φ resistance FF, and the curve resistance FFrc. A, B, C coefficient, r system curve radius.

Fr = : Frg + Fra + Frc (17 ) Frg = s (18 ) Fra = A + Bv + Cv2 (19 ) F rc = 800 / r (20 ) For the braking force calculation unit 2 1 1, the automatic operation control unit 208 The train speed V and the braking command 値Fs are input, and the data storage unit 20 1 inputs the train weight Μ and the train resistance 値Fr. The braking force calculation unit 2 1 1 calculates the braking force 値F b . using the train speed v -45 - 1284605 (42) -, the train weight Μ, and the train resistance 値F r using Equation 2 1 . The braking force 値Fb calculated by the braking force calculation unit 211 is output to the data storage unit 2 0 1 and stored. - Re-implement the instructions using the above formula. In Formula 21, the brake force is

Fb, the weight is Μ, the acceleration is α, and the train resistance 値 is Fr.

Fb=Ma + Fr (21)

The braking force calculation unit 2 1 1 calculates the deviation Fh of the braking force 値Fb calculated by the braking force calculating unit 2 1 1 and the braking command 値Fs as the braking force deviation detecting means (refer to Formula 22). The deviation Fh between the braking force 値Fb and the braking command 値Fs calculated by the braking calculation unit 211 is output to the storage unit 201 and stored in the storage unit 201. A new braking force command 値Fs can be obtained by adding the deviation Fh of the braking force 値Fb calculated by the braking force calculating unit 2 1 1 and the braking command 値Fs to the braking command 値Fs when the deviation Fh is detected. The φ calculation method can provide a more correct braking force 値Fb for the train. In Formula 22, the vehicle braking system Fb, the braking command system Fs, and the deviation system Fh. (twenty two )

Fh = Fs - Fb When the vehicle is braking, the delay time calculation unit inputs the data of the time T1 at which the automatic operation control unit 208 outputs the braking command 値Fs and the time T2 at which the train speed is decelerated. The delay time calculation unit 2 1 1 calculates the data of the time T1 at which the brake command 値 Fs is output and the data (43) 1284605 of the time T2 at which the train speed is decelerated (see Equation 23). The deviation Th calculated by the delay time calculation unit 211 is output to the data storage unit 201 and stored. The delay time Th is the time from when the actual brake command from the automatic operation control unit 20 8 is received to the brake command to reach the drive unit 205 and the brake unit 206 and the operation is performed. By detecting the delay time Th, a driving plan can be drawn up in consideration of the delay time Th, and a more correct and safer driving plan can be obtained. In the formula 23, the timing at which the automatic operation control unit 208 outputs the brake command 値F is T1 φ, the time at which the train speed is decelerated is T2, and the delay time is Th. (twenty three )

Th = T2 - The T1 data storage unit 201 inputs the train weight Mk at the time of the empty vehicle, the train weight 现时 at the current point, the number N of passengers at the time of full vehicle, and the average weight Me of the human. The boarding rate calculation unit 213 calculates the boarding rate estimation 値 Mrate using Equation 24 using the train weight Mk at the time of the empty train, the train weight 现时 at the current point, the number of passengers N φ at the time of full vehicle, and the average weight Me of the human being. The boarding rate calculation 値Mrate calculated by the boarding rate calculation unit 213 is input to the material storage unit 201 and stored in the material storage unit 201. In Equation 24, the weight of the train at the time of the empty train is Mk, the weight of the train at the current point is Μ, the number of passengers at the time of full vehicle is Ν, the average weight of the human being is Me, and the calculation of the ride rate is Mr ate. M-Mk

Mrate = —Μ〇._ (24) Ν -47- 1284605 (44)

♦ In the train characteristic learning device 207 having the above configuration, the train weight calculation unit 209 can calculate the train weight 在 during the train operation, and output the train weight of the current point to the ride rate calculation unit via the data storage unit 20 1 . Hey. Therefore, the ride rate Mrate between stations can be estimated. Since it is possible to estimate the ride rate Mrate between stations, it is possible to analyze the change in the ride rate of each station and the change in the ride rate of time. Further, since the train weight calculating unit 209 can calculate the train weight 现时 at the current point, the correct data of the train resistance 値Fr and the slope resistance 値Frg are also calculated. In the aspect of the automatic operation control unit 20, as shown in Japanese Patent Laid-Open No. Hei 5 - 1 93 5 02 and Japanese Patent Laid-Open No. Hei 6-2845-19, the position of the train is detected by the ground, the train speed, and the elapsed time, and is automatically Train operation mode (Refer to Fig. 21 (the vertical axis is the speed and the horizontal axis is the distance (position)). The target speed is determined. The train follows the target speed to implement automatic train operation control. Further, a method of detecting the position by the driving distance and the ground can be used, and the control method of the automatic operation control unit is not particularly limited. The operation control unit 208 of the present embodiment has a delay time compensation means, a running traction force deviation compensation means, and a braking force deviation compensation means which are not provided by the conventional automatic operation control unit. The delay time calculation unit 212 inputs the delay time to the delay time compensation unit (not shown) of the delay time compensation means. The delay time compensation unit (not shown) calculates the starting time of the braking force or the running traction and determines the starting time of the running traction when considering the delay time. The train weight calculation unit 209 that operates the traction deviation detecting means inputs the running traction deviation to the running traction deviation compensation means (not shown). The running traction deviation compensation means (not shown) will calculate the new -48-1284605 (45) and run the traction command 値 to control the running traction when considering the deviation of the running traction. The braking force calculation department will input the braking force deviation compensation 値 to the braking force deviation compensation means (not shown). The braking force deviation compensation means (not shown) will calculate the new braking force command and control the braking force in consideration of the braking force deviation compensation. In the automatic train operating device according to the first embodiment of the present invention, the train characteristic learning device 207 can collect information such as the ride rate, the train weight, the train φ resistance, and the braking force during driving, and not only before the safety automatic operation is implemented. The collection of information can also be applied to vehicles that use the information collected during driving to further correct the driving plan when the actual passenger is travelling. In the present embodiment, the train characteristic learning device 207 adopts a method of processing data in train driving, and the data processing can be processed after the train is driven. Further, in the present embodiment, although only the braking force is indicated, there is no limitation on the method of braking, including the braking level. Further, the train characteristic learning device of the present embodiment can collect data of rainy days, data of each section of φ, data of each route, and data of each station, etc., so it is not limited to performing data collection only for the route. . Fig. 22 is a block diagram showing the train construction of the automatic train running device according to each embodiment of the present invention. The train 0 has a speed detector 322 composed of a pulse generator (PG) or the like mounted on a rotating shaft of the wheel, and an above-ground sub-detector for detecting a ground (inquiry machine) provided on the track. Further, the automatic train running device 1 for inputting the train detecting speed signal and the train detecting position signal, and the driving device 305 and the braking device 306 by which the automatic train running device 1 performs control are further provided. Illustration omitted. 49- 1284605 (46) - The marked automatic train control unit (ATC) inputs the relevant ATC signal and operating conditions such as the speed limit to the automatic train running device -4. * The automatic train running device 1 includes a database 300, a stop-and-stop operation calculation circuit 304A, and an inter-station operation calculation circuit 304B. When the train detection speed signal and the train detection position signal are input to the station, the vehicle is implemented. The arithmetic circuit 304B. When the station is parked, the arithmetic circuit 3 04 A performs a specific calculation to be described later when the train 0 stops at the station, and when the inter-station φ vehicle performs the arithmetic circuit 304B, it performs a specific calculation to be described later when traveling between the stations of the train 0, or control. Next, the database 00 stores characteristic data such as route conditions (slope, curvature, etc.), vehicle conditions (restricted speed, vehicle weight, and train characteristics such as acceleration and deceleration performance), and timetable (running table). Various kinds of information. This database 3 00 can be a hard disk that is placed in the automatic train running device 1 or a 1C card that is recently developed and can be carried by the driver. Figure 23 is a block diagram showing the configuration of an automatic train running device 1 φ according to a thirteenth embodiment of the present invention. The arithmetic circuit 304A is implemented with the optimal driving plan drafting means 3 07, and the arithmetic circuit 3 04B is implemented with the driving plan recalculation means 308, the control command separating means 309, and the control. The command output means 310. Next, the data stored in the database 300 is input to both the arithmetic circuit 307A when the station is parked and the arithmetic circuit 304B is implemented during the inter-station driving, and the speed detector 302 and the ground sub-detector 3 03 Each of the detection signals and the ATC signal is input to the arithmetic circuit 304B only when it is input to the inter-station. The best driving plan drafting means 307 will be based on the information stored in the database 300 -50-28485 (47) • Various materials, so that the train 0 will run from one station to the next, and stop at the target time. The best driving plan for the target location. The "best" condition at this time can be various settings. For example, taking the driving time as the best - first, to improve energy efficiency is the highest priority, or to avoid the rapid acceleration and deceleration of the ride comfort is the highest priority. In addition, in the example of the method for holding the best driving plan related information of the optimal driving plan, the speed target corresponding to the time or distance is regarded as a control command.

The best driving plan is to develop a method for optimal driving plans. For example, a method of predicting train behavior using a mechanical train model (for example, Japanese Patent Laid-Open No. 5 - 1 93 5 02). As shown in Figure 37, the running curve, the taxiing curve, and the retrograde braking curve are predicted, and the intersection of the sliding curve and the retrograde braking curve is used as the starting point of the braking. The driving plan recalculation means 3 08 not only inputs the driving plan proposed by the optimal driving plan drafting means 307, but also inputs the train detecting speed and the train detecting position from the speed detector 302 and the ground sub-detector 03. The φ setting and the ATC signal from the ATC will perform the recalculation of the driving plan when the error of the proposed driving plan and the actual driving result reaches a certain level or more. The control command precipitation means 3 09 will calculate the acceleration command and the deceleration command for the current point of the drive device 305 and the brake device 306 according to the driving plan recalculation means 308 input driving plan, and output the same. To the control command output means 3 1 〇. The control command output means 3 10 0 outputs the acceleration command and the deceleration command input from the control command means 9 to the drive unit 305 and the brake device 306. -51 - 1284605 (48) - Next, the operation of Fig. 22 having the above configuration will be described. - Assume that train 0 stops at a certain station, and the best driving plan drafting means 3 07 will refer to the information stored in the database 300 to formulate the best-to-be-car plan until the next stop. Secondly, when the train 0 starts running, the driving plan recalculation means 3 8 will implement the best driving plan proposed by the best driving plan drafting means 3 07, and based on the speed detector 302 and the ground sub-detector. 3 03 The train detection speed and the train detection position are calculated and compared with the actual driving results. When the difference between the two (for example, the speed error of the difference between the speed target and the speed performance of the optimal driving plan) is greater than the preset At the point of a critical threshold, the recalculation of the driving plan will be performed. The difference between the two is greater than the critical state. In addition to the above-mentioned chasing phenomenon, it may happen that the ATC enters the speed limit change command because the other vehicle is parked in the forward direction. Further, the recalculation performed by the driving plan recalculation means 3 08 may be performed by considering the actual performance speed at the time of recalculation, the actual performance distance (train position), or the remaining time allowed for the inter-station driving. Next, the control command precipitating means 9 outputs a control command such as an acceleration command or a deceleration command from the driving plan recalculation means 3 08 to recalculate the driving plan, and the control command output means 3 1 0 outputs the outputted control command to the drive. Device 3 05 or brake device 306. With such calculation and control of the automatic train running device 3 04, the train 0 can stop at the target position of the next stop at the target time. Thereafter, during the stop of train 0 stopping at the next stop, the best driving plan will be further developed to the best driving plan to the next stop, execution and means 3 0 8~3 1 0 The same move -52- (49) 1284605 - made. Moreover, when the best driving plan proposed by the best driving plan 307 is set and the error of the actual driving result does not exceed a certain threshold, the driving plan recalculation method 3 0 8 will not perform the recalculation, but will directly The best driving plan is to output the optimal driving plan of the means 7 to the control command means 3 09. In the thirteenth embodiment of the above-mentioned Fig. 23, after the train 0 starts to drive according to the optimal driving plan proposed by the optimal driving plan drawing means 3 07, if the actual driving result and the driving plan have a certain degree of deviation from the driving plan, The recalculation of the driving plan by the driving φ plan 3 〇8 immediately implements the recalculation of the driving plan, which can greatly suppress the chasing phenomenon that has occurred in the past, so that the energy saving effect can be improved. FIG. 24 is an automatic embodiment of the first embodiment of the present invention. A block diagram of the train running device 1. The difference between Fig. 24 and Fig. 23 is the vehicle calculation recalculation means 308 of Fig. 23 using the cumulative error reference type driving plan recalculation means 311. The recalculation method 3 08 of the driving plan in Fig. 23, because at each time of recalculation, it is judged whether the error at that time exceeds the critical value φ値, so the recalculation with the chasing feeling is sometimes implemented due to the influence of the disturbance. . Therefore, in this embodiment, the cumulative error reference type driving plan recalculation means 31 1 performs judgment on the accumulation of an error (for example, an error accumulated in 5 minutes). In this manner, it is possible to prevent the above-described recalculation with the chasing sensation due to the influence of the disturbance. Fig. 25 is a block diagram showing the configuration of the automatic train running device 1 according to the fifteenth embodiment of the present invention. The difference between Fig. 25 and Fig. 24 is that the control command issuance means 309 and the control command output means 310 are provided with a control command compensation hand -53 - 1284605 (50) stomach segment 3 1 2 . The control command compensation means 3 1 2 has a function of determining whether the driving plan output by the driving plan recalculation means 308 and the error of the actual driving result exceeds a critical threshold, and when it is determined to be a critical value or more, the control command is • The control command that is deposited by the deposition means 9 performs compensation. The control command compensation means 3 1 2 is provided to enable the automatic train running device 1 to have a support function. That is, if the train 0 is based on the best driving plan, the means of calculating the means 3 07 or the driving plan recalculation means 3 0 8 is used to perform the actual driving, there is no problem, however, there is sometimes a large deviation from driving. The situation of the planned driving. For example, when one of the plural vehicles has an abnormality. However, in the present embodiment, the control command compensation means 3 12 can also perform the support function, and can appropriately compensate the control command to prevent the stop position and the target position of the train from being largely deviated. Further, in the configuration of Fig. 25, an example of the control command compensating means 3 1 2 is provided between the control command precipitating means 309 and the control command output means 3 1 0 of Fig. 23. Of course, the control command compensating means 3 12 It is also possible to provide a control finger φ between the deposition means 309 and the control command output means 3 10 in Fig. 24. Figure 26 is a block diagram showing the configuration of an automatic train running device 1 according to a sixteenth embodiment of the present invention. The difference between Fig. 26 and Fig. 25 is that the control command compensation means 3 1 2 of Fig. 25 employs the cumulative error reference type control command compensation means 3 1 3 . The control command compensation means 3 1 2 of Fig. 25, the control command compensation means 3 1 2 immediately controls the control command precipitation means 3 09 even if the occurrence of one driving plan and the actual driving result error is greater than the critical threshold. The instruction performs compensation, and is susceptible to interference and performs control with a chasing sensation. Therefore, in this embodiment, the cumulative error reference type control command -54-(51) 1284605 • The compensation means 3 1 3 performs a judgment on an error accumulated in a certain degree (for example, an error accumulated within 5 minutes). In this way, it is possible to prevent the above-mentioned recalculation with a chasing sensation due to the influence of the stalk. Fig. 27 is a block diagram showing the configuration of an automatic train running device 1 according to a seventeenth embodiment of the present invention. The difference between Fig. 27 and Fig. 26 is that the vehicle calculation recalculation means 3 0 8 is the cumulative error reference type driving plan recalculation means 3 1 1. Since the other configurations are the same as those in Fig. 26, detailed descriptions thereof will be omitted. Further, in this embodiment, the cumulative error of the driving plan and the actual driving result is judged by two means 3 1 1 and 3 1 3, however, the threshold used by these means in performing the cumulative error judgment may be Set to correspond to different conditions. Fig. 28 is a block diagram showing the configuration of an automatic train running device 1 according to a first embodiment of the present invention. The difference between Fig. 28 and Fig. 27 is that the optimal driving plan drafting means 307 for implementing the arithmetic circuit 307A when the station is parked is the late-time consideration type optimal driving plan drafting means 3 1 4, and stored in The train characteristic data of the database φ 3 00 contains "delay time" data. In the calculation of the driving plan, the delay time of the train's response to the control command, that is, the time from the output of the control command to the time when the control command affects the actual train, requires a huge computational load to obtain the aforementioned front. There is difficulty in the speed of operation in practical use. Therefore, in the present embodiment, in addition to the delay time pre-determined in the train characteristic data stored in the database 300, the optimal driving plan is also the best driving plan for the "delay time consideration" type. Means 3 1 4, this delay will also be considered when formulating the best driving plan. In this way, you can mention -55- (52) 1284605 • The stop position accuracy of the target position of the high next stop. - Fig. 29 is a block diagram showing the configuration of an automatic train running device 1 according to a nineteenth embodiment of the present invention. The difference between Fig. 29 and Fig. 28 is the cumulative error of the reference model of Fig. 28 recalculation means 3 1 1 is the delay time consideration type driving plan recalculation means 3 1 5. The delay time consideration type driving plan recalculation means 3 1 5 is the same as the delay time considering type optimal driving plan drafting means 3 1 4, and the delay time data included in the train characteristic data of the database 300 is implemented. Recalculation of the driving plan. In this way, the stop accuracy of the target position of the next parking station can be further improved. Further, in the configuration of the nineteenth embodiment, the combination of the driving schedule recalculation means 3 1 5 of the "delay time consideration type" and the optimal driving plan drawing means 3 1 4 of the "delay time consideration type" is adopted. However, it is also possible to construct a combination of the ordinary best driving plan for the non-"delay time consideration" type of means 3 07, that is, to recalculate the driving plan of the 23rd to 27th drawings. 8, 3 1 1 is replaced by the delay time considering the type of driving plan φ recalculation means 3 1 5 composition. Figure 30 is a block diagram showing the configuration of an automatic train running device i according to a twentieth embodiment of the present invention. The difference between Fig. 30 and Fig. 29 is the deferred time considering the best driving plan for drawing in Fig. 29. 3 4 4 is the best predictive driving plan for the forward forecasting type 3 i 6. Previously, the proposed method for predicting the best driving plan is also a kind of "delay time consideration type", which is based on the prediction of the direction of travel of the train to make the train stop at the target position of the next stop. For the purpose of the driving plan. That is, as shown in Fig. 38, the train behavior for calculating the direction of travel of the train -56-(53) 1284605 • Predict and perform the convergence operation at the target speed through the target point (or the progressive convergence from the start point of the deceleration) Operation), you can draw up the best driving plan without using the retrograde curve~. If you do not need to consider the delay time, you only need to refer to the target position braking characteristics and use the reverse thrust point as the starting point of the braking. The calculation will be easier. However, if the delay time must be considered, then the reverse pushing method is used. The operation will be very complicated. Therefore, it takes a lot of calculation time to obtain the start point of the brake, and the φ point may have passed the target position when the operation result of the brake start point is obtained. Further, the method shown in Fig. 3 is to perform the prediction operation of the plurality of traveling directions to obtain the starting point of the braking. Even if the calculation is performed plural times, since it can be implemented in each specific sampling period, it is only necessary to perform the comparison. Short time. Fig. 3 is a block diagram showing the configuration of an automatic train running device 1 according to a second embodiment of the present invention. The difference between Fig. 31 and Fig. 29 is the recalculation means of the delay time consideration type driving plan in Fig. 29, which is the recalculation means 3 1 7 of the forward prediction type driving plan. Previously, the recalculation φ calculation means 317 and the forward prediction type optimal driving plan preparation means 316 are the same as the forward-predicted driving plan recalculation, and the recalculation of the driving plan is implemented based on the prediction of the traveling direction of the train 0. The calculation for the purpose of stopping the train stop at the target position of the next stop station. Therefore, the recalculation of the vehicle plan considering the delay time can be implemented in a short time. In addition, the recalculation means 3 17 can be replaced with the delay time consideration type driving plan recalculation means 315 instead of the 23rd to 27th and 30th drawings. The driving plan recalculation means 3 08, 311, 315. Figure 3 is an automatic train running device 1 according to a twenty-second embodiment of the present invention

-57- (54) 1284605 • The block diagram of the composition. The difference between Figure 3 2 and Figure 3 is before Figure 31 - Recalculation of the predictive driving plan 3 1 7 is the successive forward prediction type of vehicle planning recalculation means 3 1 8 . Figure 3: Pre-predictive driving plan • Recalculation means 3 1 7 The recalculation of the driving plan is performed by performing the forward prediction operation according to each predetermined control cycle set. However, this embodiment is successively performed. The recalculation means to the predictive driving plan 3 1 8 does not have to be recalculated in each control cycle. For example, when the sampling control period is 0.3 seconds, it can be performed every 1 second, or even every 10 seconds. In this way, changing the recalculation cycle can reduce the computational load. Further, the calculation cycle can be appropriately determined in consideration of the place where the line gradient changes rapidly, and the place where the speed change is restricted. Figure 33 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty-third embodiment of the present invention. The difference between Fig. 33 and Fig. 32 is the recalculation means of the successive forward prediction type driving plan in Fig. 32. The first method is the speed measurement driving type successive forward prediction type driving plan recalculation means 3 1 9 . That is, if the detection sampling period of the φ speed detector 312 is 1 [msec], the operation circuit 304B side of the inter-station driving operation is not directly using the speed detection signal input according to the period, but is 5 to 1 0 [msec] During the period, the speed detection signal input is subjected to filtering and the like, and then the data update is performed. Secondly, the speed measurement drive type progressive forward predictive driving plan recalculation means 3 1 9 is based on the update cycle of this data to implement the recalculation of the forward predictive driving plan. In this way, the influence of interference and the like can be suppressed, and the arithmetic precision at the time of recalculation can be improved. Figure 34 is a block diagram showing the configuration of an automatic train running device -58-(55) 1284605 •10 in the twenty-fourth embodiment of the present invention. In this embodiment, in addition to the inter-station driving calculation circuit 309B of the inter-station driving station in FIG. 3, an inter-station driving result storage means 320 is added, and "the delay time estimation means 2 is added to the arithmetic circuit 304A when the station is parked." 1, and can calculate the delay time based on the latest driving results. Therefore, the database 3 of this embodiment may not store the delay time data. That is, after the train 0 departs from a certain station, the inter-station traffic φ fruit data during the period from the train position, the train speed, the ATC signal, and the like to the next stop station is stored in the inter-station driving result storage means 320. . Secondly, after the train arrives at the next stop and stops, during the stop, the delay time estimating means 321 calculates the delay time based on the data stored in the inter-station driving result storage means 3 20, and outputs the estimated result to the delay time. The best driving plan drafting means 314 and the forward predictive driving plan recalculation means 3 17 . The delay time consideration type optimal driving plan drafting means 314 and the forward predictive type driving plan recalculation means 3 17 will further implement the line to the next stop station in consideration of the estimated delay time. The formulation and recalculation of the φ car plan. If the method of estimating the delay time by the delay time estimating means 321 is explained, this method does not use a complicated operation, but is a simple method of estimating the signal level change based on the measurement data. For example, when the vehicle is braked, after the brake control command is output and the level operation is performed, the train speed decreases after a certain period of time elapses. At this time, the time until the threshold 値 which is set in advance is estimated to be a delay time. Further, the delay time stored in the database 300 of the above-described 28th to 33rd drawings is particularly complicated because it is not required to be obtained in a time-limited state, so that the complex -59-(56) 1284605 can be used. Calculate and store the result of the calculation, and carry out the test driving of the train 0. It is easier to obtain the data by using the method of estimating the delay time 321 . This embodiment is delayed due to the delay of reflecting the latest train characteristics. Time-considered optimal driving plan development method 3 1 4 and forward-predicted driving plan recalculation means 3 1 7 The proposed and recalculated driving plan can further improve the reliability.

Figure 35 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty-fifth embodiment of the present invention. The difference between Fig. 5 and Fig. 34 is to implement the additional line delay time derivation means 322 on the arithmetic circuit 3 0 4B when driving between stations, and the forward predictive type driving plan recalculation means 3 17 can be considered. In the case where the delay time estimated by the online delay time estimating means 22 is calculated, the recalculation is performed. That is, in the composition of Fig. 34, the delay time is calculated based on the inter-station driving result of a certain section, and the result of the calculation is applied to the recalculation of the φ car plan of the next section, this 3 5 In the embodiment of the figure, even if it is driving in the same section, the delay time can be estimated based on a small number of inter-station driving results, so it can also be applied to recalculation. Therefore, the recalculation result of the predictive driving plan recalculation means 3 1 7 before this embodiment is more representative of the latest train characteristics than the one shown in Fig. 34. Figure 36 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty sixth embodiment of the present invention. In the embodiment, the arithmetic circuit 3 04B is added to the interim predictive parking temporary parking plan calculation means 3 23 and the driving plan adoption means 3 24 when the vehicle is traveling between the stations in Fig. 5 . Next, in this embodiment, 8 - 60 - (57) 1284605 - is divided into three types of PI, P2, and P3 in accordance with the train driving time, and when the train 0 reaches a specific point on the front side of the target position, the driving meter The painting 'using means 324 will use the forward-predicted parking temporary driving plan calculations. Means 3 23 Calculate the driving plan P3. Hereinafter, the twenty-sixth embodiment will be described in detail. First, the definitions of the driving plans PI, P2, and P3 are as follows.

P 1 : When the train 1 stops at the station, the best driving plan proposed by the driving plan 3 1 4 (or 307, 316) can be calculated by the driving plan. P2: In the inter-station driving of the train 1, the recalculation method is implemented by the recalculation method of the driving plan 3 17 (or 308, 311, 315, 318, 319) to recalculate the driving plan. P3: The time of the N-meter (for example, N = 3 00 [m]) point before the train arrives at the train station and the train 0 reaches the target position. Proposed temporary parking plan for parking. When the train 〇 reaches the target position N meters before, the temporary driving plan calculation means 323 formulates the subsequent parking temporary driving plan P3 at a specific cycle (for example, the detection sampling cycle of the speed detector 2). This parking use temporary travel plan P3 is designed to use the train detection speed and train detection position at that time to predict the train's parking behavior in consideration of the delay time of the train travel direction. This parking behavior is, for example, pre-planning a basic parking behavior when the train is stopped at a specific braking level position at the current point, and is used to stop. Secondly, in terms of train driving prediction, the physical model of the following formula (25) can also be considered.

-61 - (58) 1284605 F - Fr = M · a ( 25 ) F: Running traction or braking force

Fr : train resistance (driving resistance, slope resistance, curve resistance, tunnel resistance, etc.) Μ : train mass α : acceleration or deceleration

The resistance of the train resistance Fr series when driving, for the convenience of calculation, as described above, the composition of driving resistance, slope resistance, curve resistance, and tunnel resistance are usually considered. Therefore, the train resistance Fr can be obtained by the equation (26).

Fr = Frg + Fra + Frc + Frt (26) Each of the resistances in equation (26) is obtained using the φ data stored in the database 300 using the following resistance equations (27) to (30) (see " Operation theory (DC AC power car), edited by the Dating Society. • Slope resistance

Frg = s ( 27 )

Frg: slope resistance (kg weight / ton) S: slope (% 〇) (positive on the uphill slope, negative on the downhill slope) • driving resistance

Fra=A+B*v + C.v2(square of v) (28)

Fra: driving resistance (kg weight / ton)

-62- 1284605 (59) - A, B, C: Coefficient • v: speed (km/h) "•curve resistance w Frc = 8 00/r ... (29)

Frc: curvature resistance (kg weight / ton) r: curve radius [m]

• Tunnel resistance type (Because the tunnel resistance will vary greatly depending on the shape and size of the tunnel section and the speed of the train, etc., the following may be used for convenience)

Frt= 2 (for single-line tunnel) or =1 (for double-track tunnel) (30)

Frt: tunnel resistance (kg weight / ton) Temporary driving plan calculation means 3 23 Because the physical model of the above formula (25) is used, after the N-meter position is reached before the target position, the parking temporary will be repeated. Driving plan P3. By repeating the plan, the parking position of the temporary parking plan for parking P3 is gradually approached to the target position. As shown in Figure 39. Further, the distance N from the target position to the start position of the temporary parking plan for parking can be determined by the formula "travel distance" ± "wide distance". Next, the operation of the driving plan employing means 324 of Fig. 3 will be described with reference to the flowchart of Fig. 40. When a driving plan for one of PI, P2, and P3 is formulated or recalculated according to a specific cycle, the flow chart is a processing step of one cycle.

-63- (60) 1284605 • First of all, the 'vehicle plan adopts means 324 to judge the current train 0. The state or the time when driving is stopped by the station or just after the departure from the station, the station line, the time of the vehicle, and whether it is located. Near the target parking location (step 1). Secondly, when β is judged as “After the stop at the station or just after the departure from the station”, the optimal driving plan 拟 1 (step 2) proposed by the late-time optimal driving plan drafting means 314 is adopted. Thereafter, the driving plan employs means 3 24 to output the optimal driving plan Ρ 1 to the control command discharging means 3 09. Further, the operation after the control command φ deposition means 309 inputs the driving plan has been described in the above embodiment, and thus the overlapping description will be omitted. When it is judged at the time of "the inter-station driving" in step 1, the driving plan employer means 324 judges whether or not the recalculation of the driving plan of this cycle has been carried out (step 3). Next, if the recalculation has been carried out, the forward planning vehicle recalculation means 317 recalculates the recalculation of the driving plan Ρ 2 (step 4). On the other hand, if it is judged in step 3 that the recalculation of the driving meter φ drawing of this cycle is not performed, it is judged whether or not the first driving cycle Ρ 1 has been adopted in the previous cycle, that is, the previous cycle (step 5). If the best driving plan Ρ1 has been taken at the first hour, the driving plan 324 will use the best driving plan Ρ1 (step 2). However, when the best driving plan 未1 was not used at the first 1 o'clock, the current point is the best driving plan. 1 The time has been adopted and the recalculation has been carried out. The first 1 hour point is the recalculated driving. plan. Therefore, the driving plan uses the means 324 to adopt the driving plan adopted at 1 o'clock (step 6), and the judgment of step 1 is "near the target parking position", that is, -64-1284605 (61) • target parking position When the meter is less than N meters, the means of driving the vehicle will be used. 3 2 4 Yes • Enter the temporary parking lot that has been prepared by the temporary driving plan calculation method 3 23, and determine whether the parking position is at the target parking position. • Within ± within the tolerance range (step 7). Next, if the parking position is within this range, the parking temporary driving plan P3 is adopted (step 8). However, if it is not within this range, then return to step 5, and use the recalculation of the driving plan at the first 1 o'clock (or before), and then repeat the step 7 after the step 1 is repeated. Until it is within range. As described above, in the twenty-sixth embodiment, the parking temporary parking plan in which the train is stopped in the "target parking position" ± "permissible error" near the target parking position is used, and the train can be stopped at the target with good precision. Parking location. Further, since the train is scheduled to move in the traveling direction, the temporary parking plan for parking is proposed, and it is easy to obtain an automatic train running device which is very convenient in considering the delay time and has a very simple calculation. In the embodiment, the parking temporary planning calculation means φ 3 23 is an example of the "forward prediction type". However, the parking temporary driving calculation means 323 is not limited to "Forward predictive type". The automatic train running device of each of the embodiments described so far is a method of changing the control command in stages by the running level and the braking level of the current general train. However, in the near future, the command signal should be continuously controlled to drive the drive unit and the brake unit. Therefore, as long as the control command for acceleration is made into a continuous traction command or a running torque command, it is better to implement the optimal driving plan or the driving plan to re--65- 1284605 (62) 1 'Automatic operation with comfort and higher energy savings. In addition, the control command during deceleration can be used as a continuous braking force command, and the optimal driving plan or the recalculation of the driving plan can be implemented, and the automatic riding comfort and the energy saving effect can be realized automatically. Running. Alternatively, both of the above-described continuous control commands during acceleration and deceleration can further achieve automatic operation with better ride comfort and higher energy saving effects.

Next, a twenty-seventh embodiment will be described with reference to the drawings. Figure 41 is a schematic diagram showing the configuration of an embodiment of the present invention. The speed position calculation unit 405 calculates the speed and position of the train 0 in the train based on the information of the speed detecting unit 403 such as the tachometer and the information of the ground detecting unit 404 that detects the signal of the ground. The data acquisition means 4 1 2 now inputs it to the train position stop automatic control device 410. Further, the map is not indicated, and the information such as the brake level and the stop target position is also input to the train position stop automatic control device 4 1 0 via the train current data acquisition means 4 1 2 . The train position stop automatic control device 4 1 0 is based on the current speed, the current position, the current brake level, and the like obtained by the train current data acquisition means 4 1 2, and the data stored in the brake characteristic data storage unit 41 1 . The vehicle characteristic data such as the deceleration of the brake class, the delay time of the brake class switching, and the response delay time, using the deceleration control plan drafting means 4 1 3 to formulate the deceleration control in which the combination of the complex levels is used to stop the train at the stop target position plan. For example, when the train is stopped at a specific position in a combination of two levels, the deceleration control plan calculates the time allocation of each brake class. First, the first-66-(63) 1284605 •1 brake level is maintained in the aforementioned time allocation calculation station. After the specified time has elapsed, • switch to the second brake level and maintain until the train stops. Figure 42 is the simplest example of a reduced speed control plan. This example is a deceleration control plan for the remaining distance l〇m _ point, and the level is switched around the remaining distance of 6 m to stop the train at the target stop position. For the time allocation, for example, the fake design drawing uses two levels, and the maintenance time of the first braking level is regarded as a variable for the current speed and the remaining distance, the driving distance when decelerating with the first braking level, and the second braking level of the φ second braking level. The total distance of the driving distance during deceleration is equal to the remaining distance mode, and the maintenance time of the first braking level can be obtained, thereby obtaining the time allocation. If there is no solution that satisfies the condition, the combination of the two levels can be changed and the same calculation can be repeated. In the calculation of the driving distance, it is assumed that the deceleration of the braking level before the switching is performed during the period of the delay switching time after the braking level output command, and the response delay time after the delay time elapses from before the switching. The deceleration of the brake class gradually changes to the deceleration of the brake class after the switch. After the response delay time elapses, the deceleration is performed with the deceleration of the brake class after the φ switch, and the calculation of the temporary travel distance is implemented under the above assumption. Plans to consider the braking response characteristics when considering level switching. The deceleration of each brake class is maintained at a safe time. According to the plan switching level proposed in this way, the train can be stopped at a specific position without frequent switching of the level. In addition, when planning the plan, the first brake level is a level with a large deceleration, and the second brake level is a level with a small deceleration. When the vehicle is parked at a lower level, the ride comfort can be improved. When the deceleration rate of each brake class is changed, for example, when the maintenance time of the first brake level (the level of the large deceleration) is exceeded (the switching schedule time -67-(64) 1284605 -), the plan will be adopted. The predicted speed at which the deceleration is decelerated is compared with the actual train speed. If the actual speed is small, that is, the deceleration is less than the assumed hour, do not immediately switch to the second brake level (the deceleration is small). Level), use the extension of the first brake class to prevent the train

Exceeded the target stop position. Fig. 43 is an example of adjusting the stop position by changing the switching schedule time. In this example, the actual deceleration is less than the hypothesis, and the deceleration is slower. Therefore, the initial plan is scheduled to be switched between 5m and the deceleration is changed to a level of 3.2m to switch, and the stop position is adjusted. Fig. 44 is a flow chart for adjusting the stop position by changing the switching timing. To extend the maintenance time, for example, to calculate the actual deceleration according to the actual train speed at the switching plan time, and to calculate the deceleration control plan at the time when the first brake level command is recalculated by the estimated deceleration, or according to the calculation Deceleration, recalculate the plan to start the switching plan. Also, when formulating the initial deceleration control plan, the maximum preset deceleration is used, and the level switching time can be extended to adjust the stop position regardless of whether the actual deceleration is small or large. Figure 45 is a schematic block diagram of a twenty-eighth embodiment of the present invention. The rest of the configuration is the same as that of the twenty-seventh embodiment except that the deceleration estimating means 416 for estimating the deceleration based on the time series data of the train speed during deceleration is used, and the basic functions are also the same. The deceleration estimation using the deceleration estimating means 4 1 6 can be obtained by, for example, a specific time after a delay of the level switching and a response delay time elapsed at a specific deceleration corresponding to the level. The speed should be slowed down to calculate the deceleration. Train speed

-68 - (65) 1284605 • When there is a large error, the moving average of the speed should be taken, and the deceleration should be estimated based on the data after the interference is removed by appropriate filtering. The deceleration estimating means 4 1 6 is used to estimate the deceleration of the time point, and the deceleration control plan is corrected by the deceleration obtained by the calculation, so that the deceleration of each braking level is caused by the time or speed in one driving. When changing, you can also obtain a correspondence to ensure the stop accuracy. Figure 46 is a schematic block diagram of a twenty-ninth embodiment of the present invention. The rest of the configuration is the same as that of the twenty-seventh embodiment, and the basic functions are the same, and the plan deceleration correcting means 417 performs the time points or the respective times when the deceleration control plan is decelerated. The comparison between the predicted speed of the position and the actual train speed, and the deceleration used for the difference correction deceleration control plan. According to the calculation of the predicted speed at each time point or each position when the deceleration control plan is decelerating, for example, after calculating the brake level used for the plan and the time allocation, the brake level used according to the current train speed and the plan is used. The deceleration, the level switching delay time, and the response delay time are calculated. The predicted speed can be stored in the array mode from the start to the stop of the project, or can be referred to successively. If the memory capacity of the control computer is limited, the train speed of the previous time can also be used. The deceleration of the brake class is calculated successively. When comparing the predicted speed at this time and the actual train speed, when the train speed is small, the actual deceleration should be greater than the deceleration used in the plan. Therefore, the deceleration should be increased and the deceleration control plan should be recalculated. Conversely, when the train speed is large, the actual deceleration should be less than the deceleration used by the plan. Therefore, the deceleration should be reduced and the -69- (66) 1284605 deceleration control plan should be recalculated. When the deceleration is changed, for example, the error of the predicted speed and the actual train speed is set, and the amount of change in the deceleration is determined corresponding to the time when the error tolerance is reached. By using the plan deceleration correction means 4 1 7 -, the comparison of the predicted speed and the actual train speed is performed successively and the deceleration is corrected, and the deceleration control meter can be appropriately updated corresponding to the time change of the deceleration. Since there is an error in the data of the actual train speed, it is better to use the filtered data or set the upper and lower limits of the deceleration change amount.

[Effect of the Invention] In the present invention, in addition to ensuring that the train is stopped at a specific stop position at a specific time, it is possible to realize an energy-saving operation for reducing energy loss caused by driving. Further, the present invention can implement on-line train characteristics, route characteristics, and automatic learning of control parameters, and use the learning result to realize automatic train operation with an effective φ rate. Further, the present invention can provide a device for collecting the necessary information for operating the operation device when the train travels to and from the scheduled route. Further, the present invention achieves an energy saving effect by eliminating the influence of chasing during automatic operation of the train. Further, according to the specific embodiment, it is possible to improve the stop accuracy of stopping the train at the target position by obtaining the delay time, and in other embodiments, it is possible to improve the poor ride feeling due to the change in the stage of the speed control command during the level operation. -70- (67) 1284605 • In addition, the present invention is based on the vehicle characteristic data such as the deceleration of each train level of the train, the delay time of the brake/level switching, and the response delay time, • the current speed of the train, the current position, and now The data of the brake class, etc., is intended to be a deceleration control plan for the purpose of stopping the train at a specific position by using a plurality of brake classes. Further, even if the deceleration can only be set by discrete turns, it is possible to switch the ranks frequently. In this case, it is also possible to formulate a plan for stopping the train at a specific position, and according to the plan, the ride comfort of the φ during the deceleration control is improved and the stop accuracy is ensured. Further, the present invention uses a combination of a plurality of brake classes to calculate a time distribution of each brake class for the purpose of stopping the train at a specific position, and constructs a deceleration control by the switching timing of the used brake level and the brake level. In this way, when the deceleration is changed, the time distribution can be changed, and the stop position can be adjusted without changing the level of the movement, thereby improving the ride comfort and ensuring the stop accuracy. Further, in the deceleration control plan of the present invention, the deceleration is first performed at a 煞φ car level with a large deceleration, and then the vehicle is switched to a lower deceleration level, and the parking is performed at a reduced deceleration level, thereby improving In the comfort of the ride, the present invention implements a comparison between the predicted speed of the switching time at the time of deceleration and the actual train speed at the time of switching according to the deceleration control plan, and the deceleration control plan is changed when the two are different. In this way, it is easy to evaluate the actual train deceleration condition, and the deceleration control plan corresponding to the change of the deceleration can be recalculated to improve the stop accuracy. Also, after the proposed deceleration control plan was drafted,

-71 - (68) 1284605 • The deceleration control plan can be changed when the plan is used differently. This method can improve the ROUBUST control for the control of the deceleration disturbance and ensure the stop accuracy. Moreover, the present invention estimates the deceleration according to the time series data of the train speed during deceleration, and formulates the deceleration control plan according to the estimated deceleration speed. With this method, the ROUBUST generation for controlling the deceleration variation interference can be improved, and Ensure stop accuracy without cumbersome adjustments. Further, the present invention performs a comparison between the predicted time at each time point or each position when the deceleration control program is decelerated, and the actual train speed, and the deceleration used in the difference correction deceleration control plan, and the deceleration according to the correction. By changing the deceleration control plan, this method can improve the ROUBUST performance for the control of the deceleration fluctuation disturbance, and ensure the stop accuracy without complicated adjustment. Moreover, according to the speed of the previous time step, the deceleration used in the proposed plan, the level switching delay time, and the response delay time, the present invention calculates the time points or positions at the time of deceleration according to the deceleration control plan. In this way, when the memory capacity of the control computer is limited, the ROUBUST property of the control for the deceleration fluctuation disturbance can be improved, and the stop accuracy can be ensured without complicated adjustment. BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram of an automatic train operating device according to a first embodiment of the present invention. Figure 2 shows the machine loss index and the total loss indicator at runtime.

-72- (69) (69)

1284605 _ example map. Figure 3 is an example of machine loss indicators, brake loss indicators, and total loss indicators for braking operations. Figure 4 is an example diagram of the converter loss indicator and motor loss indicator during operation. Figure 5 is an example diagram of converter losses and motor losses during operation. Fig. 6 is a view showing an example of the driving mode of the first embodiment. Fig. 7 is a block diagram showing an automatic train operating device according to a second embodiment of the present invention. Fig. 8 is a block diagram showing the brake loss when the running load is limited. Fig. 9 is a block diagram showing the automatic train operating device according to the third embodiment of the invention. Fig. 10 is a block diagram showing a train operation support device according to a fourth embodiment of the present invention. Fig. 1 is a block diagram showing a configuration example of a thrust indicating device according to a fourth embodiment.

Fig. 12 is a block diagram showing the control system of the thrust indicating device of Fig. 11. Fig. 3 is a block diagram showing a configuration example of a thrust indicating device of the train operation support device according to the fifth embodiment of the present invention. Fig. 14 is a block diagram showing a configuration example of a thrust indicating device of the train operation support device according to the sixth embodiment of the present invention. Fig. 15 is a block diagram showing the entire train having the automatic train running device of the present invention. (§) -73- 1284605 (70) " Figure 16 is an illustration of the internal structure of the automatic train running device in Figure 15. • Block diagram. Figure 17 is a diagram of the driving mode compensation calculated based on the weight of the initial operation. Figure 18 is a flow chart showing the steps of learning before and after business. Fig. 19 is a block diagram of a compensation means for the purpose of compensating for automatic characteristic learning results according to an embodiment of the present invention.

Figure 20 is a block diagram of an automatic train running device and a data storage unit. Figure 21 is an example of an automatic train operation mode. Fig. 22 is a block diagram showing the arrangement of trains of the automatic train running device according to each embodiment of the present invention. Figure 23 is a block diagram showing the configuration of an automatic train running device 1 according to a thirteenth embodiment of the present invention. Fig. 24 is a block diagram showing the configuration of an automatic train running device 1 according to a fourteenth embodiment of the present invention. Figure 25 is a block diagram showing the configuration of an automatic train running device 1 according to a fifth embodiment of the present invention. Figure 26 is a block diagram showing the configuration of an automatic train running device 1 according to a sixteenth embodiment of the present invention. Figure 27 is a block diagram showing the configuration of an automatic train running device 1 according to a seventeenth embodiment of the present invention. Figure 28 is an automatic train running device of the eighth embodiment of the present invention! The composition of the block diagram. Moving train operating device 1 is a second embodiment of the present invention.

-74- (71) 1284605 1 scoop to form a block diagram. Fig. 30 is a block diagram showing the configuration of an automatic train running device 1 according to a twentieth embodiment of the present invention. Fig. 3 is a block diagram showing the configuration of an automatic train running device 1 according to a second embodiment of the present invention. Fig. 3 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty-second embodiment of the present invention.

Fig. 3 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty-third embodiment of the present invention. Fig. 4 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty-fourth embodiment of the present invention. Figure 35 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty-fifth embodiment of the present invention. Figure 36 is a block diagram showing the configuration of an automatic train running device 1 according to a twenty sixth embodiment of the present invention. Fig. 3 is a diagram showing an example of the characteristics of the optimum driving plan to be proposed in the embodiment of the present invention. Fig. 3 is a diagram showing an example of characteristics of a driving plan to be formulated or recalculated in the embodiment of the present invention. Fig. 39 is a diagram showing an example of the characteristics of the temporary driving plan proposed in the embodiment of the present invention. Fig. 40 shows the flow chart of the action plan of the means 24 in the driving plan of Fig. 36. FIG. 4 is a diagram of the train position stop automatic control device of the present invention.

-75- (72) 1284605 A schematic diagram of the application of the example. > Fig. 42 is a schematic diagram showing an example of a deceleration control plan employed by the train position stop automatic control device of the present invention. Fig. 43 is a schematic diagram showing an example of changing the stop position by changing the switching schedule of the train position stop automatic control device of the present invention. Figure 44 is a diagram showing an example of changing the stop position adjustment step of the switching schedule adjustment stop position of the train position stop automatic control device of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 45 is a schematic block diagram showing a twenty-eighth embodiment of the train position stop automatic control device of the present invention. Fig. 46 is a schematic block diagram showing the twenty-ninth embodiment of the train position stop automatic control device of the present invention. Fig. 47 is a block diagram showing a configuration of a general train system having an automatic train running device. Figure 48 is a block diagram of the automatic train running device of the system of Figure 47.

[Description of component symbols] 0: Train 1: Automatic train running device (ΑΤΟ) 2: Drive brake device 3: Database 4: VVVF inverter transformer 5: Main motor 6: Brake control device (§) -76 - 1284605 (73) Tractor car inspection and inspection wheel mechanical speed Μ car inspection and inspection on the roadway 678901234 111122222 car order type standard deviation test line finger finger finger interface distance transport When the Jiali car is out of the car, the car is the most popular. The car is the most popular. The main line is the main entrance. The mouth is painted. The livelihood is calculated. The tlmll is used to calculate the 4BP Ε 2 ΠΤ one by one. Forcing 70 7VJ 立口立口重重部&H &ΗΠ CHM. Supplemental dismounting aids β. 咅β. Ε β, 咅 咅 令 制 制 制 指控 指控 指控 指控 指控 指控 指控The display table is up to the level of the horse and other services and the code is discussed and edited. -77- 1284605 (74) 31: Light 3 2: Suggested level indicates the control unit 3 3 : Sound output unit 3 4 : Database 3 5 : Driving mode precipitation unit 3 6 : database 102: automatic train control device (ATC)

103: database (DB) 104: driver's station 105: load device 106: speed detector 107: above ground detector 1 0 9 : drive device 1 1 0 : speed reduction device 120: pre-business driving determination means 1 2 1 : Pre-service characteristic initial setting means 1 22: pre-operating test train automatic operation means 123: driving result storage means 124: pre-service characteristic estimation means 125: estimation result compensation means 126: characteristic estimation 値 storage means 1 3 0 : Learning characteristic database (learning characteristic DB) 1 3 1 : Characteristic initial setting means 132: Automatic train operation means -78-1284605 (75) 1 3 3 : Driving result storage means 134 after business: Post-business characteristic learning means 1 3 5: Learning result compensation means 1 3 6 : Learning result comparison means 1 3 7 : Learning result compensation means 1 8 0 : Data processing means 1 8 1 : Train automatic operation means

1341 to 1 345 : automatic characteristic learning means 2 0 1 : data storage unit 203 : above-ground sub-detector 204 : speed detector 205 : drive device 2 0 6 : brake device 207 : train characteristic learning device 208 : automatic operation control unit 209 : Train weight calculation unit 2 1 0 : Train resistance calculation unit 2 1 1 : Brake force calculation unit 2 1 2 : Delay time calculation unit 2 1 3 : Travel rate calculation unit 300 : Database 3 02 : Speed detector 303 : Above-ground detector 3 04A: Operation circuit when stopping by station -79- (76) 1284605 3 04B: Operation circuit when running between stations 3 0 5: Drive unit 3 06: Brake device 3 07: Best driving plan Formulation means 3 08: Driving plan recalculation means 3 09: Control instruction precipitation means 3 1 0 : Control instruction output means

3 1 1 : Cumulative error reference type driving plan recalculation means 3 1 2 : Control command compensation means 3 1 3 : Cumulative error reference type control command compensation means 3 1 4 : Delay time consideration type optimal driving plan drafting means 3 1 5 : Delayed consideration of driving plan recalculation means 3 1 6 : Forward predictive optimal driving plan formulation means 3 1 7 : Forward predictive driving plan recalculation means 3 1 8 : Successive forward forecasting Type of driving plan recalculation means 3 1 9 : Speed measuring drive type progressive forward predictive driving plan recalculation means 320: inter-station driving result storing means 321 : delay time estimating means 322 : online delay time estimating means 3 23 : Forward predictive parking temporary travel plan calculation means 324: Driving plan adoption means 402: braking device 403: speed detecting unit 1284605 (77) 404: above-ground sub-detecting unit 405: speed position calculating unit

4 1 0 : Train position stop automatic control device 4 1 1 : Brake characteristic data storage unit 4 1 2 : Train current data acquisition means 4 1 3 : Deceleration control plan preparation means 4 1 4 : Deceleration control command precipitation means 4 1 5 : Deceleration control command output means 4 1 6 : Deceleration estimation means 4 1 7 : Plan deceleration correction means

Claims (1)

1284605 ΤΊΓΤΤΓ ——— Year R Repair (More) X. Patent Application No. 95111232 Patent Application Revision of Chinese Patent Application Revision of the Republic of China on February 13, 1996 1. An automatic train running device featuring: The data processing means for obtaining the information at the time of train driving is processed on-line; the data obtained during the train driving, and the information obtained in advance are obtained by using the data processing means, and the control parameters and the train when the train is driving are automatically learned when the train is driving. The automatic characteristic learning means of the characteristics and the route characteristics; and the train automatic operation means for performing the automatic operation of the train using the train characteristics and the route characteristics learned by the automatic characteristic learning means. 2. For example, the automatic train running device of the first application of the patent scope has the pre-commercial characteristics of the train characteristics and route characteristics necessary for the automatic operation of the train before the pre-business test, and the pre-business characteristics of the control parameters. The means 'the automatic characteristic learning means described above is based on the initial calculation of the pre-business characteristic estimation means, and performs the driving learning after the use of the business. 3. The automatic train running device of claim 1 or 2, wherein the automatic characteristic learning means performs learning when the driving characteristics of the train are judged to be different from the actual one, and the learning content is reflected in Then the train was on the train. 4. For example, the automatic train running device of the first or second patent application scope, 1284605, wherein the above-mentioned automatic characteristic learning means performs learning based on the driving result of one station, and reflects the learning content to the next station. Train driving. 5. The automatic train running device of claim 1 or 2, wherein the automatic characteristic learning means performs learning based on the driving result of the 1 route, and the learning content is reflected in the next route driving. 6. The automatic train running device of claim 1 or 2, wherein the automatic characteristic learning means performs learning based on the driving result on the 1st day, and the learning content is reflected in the train driving the next day. 7. The automatic train running device of claim 1 or 2, wherein the automatic characteristic learning means performs learning based on the driving result of at least several days, and the learning content is reflected in the train driving after the second time. . 8. The automatic train running device of claim 1 or 2, wherein at least two of the foregoing automatic characteristic learning means are combined; and the learning result of comparing the learning results of the automatic characteristic learning means is further provided. Comparison means; and compensation means for learning results that compensate for each of the -2- 1284605 learning outcomes based on the comparison of the learning outcome comparison means. 9. For example, the automatic train running device of claim No. 12 of the patent scope has a means for calculating the result of the calculation, and the calculation result using the pre-business characteristic estimation means is a characteristic that is practically impossible to occur, or is actually separated from the actual When the limit characteristic may occur, the compensation of the above-mentioned estimation node is performed so as to be within the aforementioned limit characteristic 値. 1 〇, as in the automatic train running device of claim 1 or 2, Φ has a second learning result compensation means, and the learning result by the above-mentioned automatic characteristic learning means is a characteristic which is practically impossible, or When deviating from the limit characteristic that may actually occur, the compensation of the learning result is implemented to be within the limit characteristic. 1 1. For the automatic train running device of claim 1 or 2, in which the automatic train running device of the automatic train operation is implemented according to the error of the target driving plan and the compensation of the control command, the aforementioned automatic learning The characteristic means implements the characteristic learning according to the control instruction compensation amount according to the error between the target and the driving plan during the execution of the characteristic learning during the driving operation. 3-1 1. The automatic application of the first or second patent range The train running device, wherein the aforementioned automatic characteristic learning means performs the characteristic learning using an adaptive observer. 2 1 3. If the automatic train running device of the patent scope 〖 or 2 is applied, 1284605, the aforementioned automatic characteristic learning means uses the interference observer to perform characteristic learning. 14. An automatic train running device, comprising: a train characteristic learning means for collecting train characteristics and route characteristic information in a train driving; and calculating a train optimum according to the train related information collected by the train characteristic learning means; The operation mode, and the automatic operation control unit that performs the automatic operation of the train according to this mode. 1 5. An automatic train running device according to claim 14 of the patent scope, wherein the train characteristic learning means and the series vehicle weight calculating means are used. 1 6. The automatic train running device of claim 14 of the patent scope, wherein the train characteristic learning means and the series vehicle resistance calculating means are used. 0 1 7. The automatic train running device of claim 14 of the patent scope, wherein the train characteristic learning means is a vehicle braking force calculation means. 18. The automatic train running device of claim 14 of the patent scope, wherein the train characteristic learning means is a delay time calculation means. 1. The automatic train running device of claim 4, wherein the train characteristic learning means is a ride rate calculation means. -4 - 1284605 2 0. The automatic train running device of claim 14 of the patent application, wherein the train characteristic learning means is a route shape calculating means. 2 1. The automatic train running device of claim 14 of the patent scope, wherein the train characteristic learning means is a slope resistance calculating means. 2 2. For the automatic train running device of claim 14 of the patent scope, the above-mentioned train characteristic learning means is a running traction deviation detecting means for detecting the deviation of the running traction command and the running traction force. 23. The automatic train running device according to claim 14 of the patent scope, wherein the train characteristic learning means is a vehicle force deviation detecting means for detecting a deviation between the braking force command and the braking force. > 24. The automatic train running device according to claim 14 of the patent application, wherein the automatic operation control unit calculates the delay time by the train characteristic learning means, and the delay time compensation means 25 for performing the delay time compensation The automatic train running device of claim 14, wherein the automatic operation control unit detects the running traction command and the running traction force by the train characteristic learning means, and compensates for the running traction command and the running traction. Deviation of the running traction deviation compensation hand -5 - 1284605 segment. 26. The automatic train running device of claim 14, wherein the automatic operation control unit detects the braking force command and the braking force when the train characteristic learning means detects the deviation of the braking force command and the braking force. The deviation of the braking force deviation compensation means. (S) -6-