JPH03163257A - Speed change stage determining device for automatic transmission - Google Patents

Abstract

PURPOSE:To prevent the generation of the improper speed change stage by excluding at least one signal system of a variety of traveling parameters when the above-described signal system fails and determining the aimed speed change stage by the fuzzy estimation or speed change map by using the rest signal systems. CONSTITUTION:A fail detecting means judges if each inputted signal system fails or not, and if a fail is generated in any signal system, the aimed speed change stage K is set to a standard speed change stage N', and the fuzzy estimation is bypassed. Accordingly, the aimed speed change stage is determined on the basis of the standard speed change stage N' which is obtained from the car speed and a speed change map for the engine load. If, in this method, a fail is detected in one system among the signal systems, fuzzy estimation is suspended at all, and the aimed speed change stage is determined only by the speed change map, and the need of rewriting a membership function related to the fail is obviated, and the control flow is excluding simplified.

Description

[Detailed description of the invention] [Industrial application field]

The present invention relates to a gear position determination device for an automatic transmission configured to determine a target gear position by performing fuzzy inference based on signals from sensors for various driving parameters.

[Conventional technology]

A gear shift mechanism and a plurality of frictional engagement devices are provided, and engagement of the frictional engagement devices is selectively switched by operating a hydraulic control device to achieve which of the plurality of gears. Stepped automatic transmissions for vehicles configured to do this are already widely known. Such automatic transmissions generally include a shift lever operated by a driver, a vehicle speed sensor that detects vehicle speed, and a throttle sensor that detects throttle opening, which is considered to reflect engine load. It is configured to selectively switch the engagement state of the friction engagement platform device according to the range of the shift lever and according to a preset shift map of vehicle speed and throttle opening. The shift map is set as shown in FIG. 28, for example. If you are currently driving at point A (overdrive gear: 4th gear) with vehicle speed n1 and throttle opening θ1, when the accelerator pedal is depressed and the throttle opening becomes θ2, the position on the map will change. The vehicle will move to point B, and the automatic transmission will be shifted to the third gear (see the broken line from 0/D to 3). By the way, the shifting of conventional automatic transmissions is based on this configuration, and the shifting point (the position of the shifting line on the map) in the vehicle speed-engine load shifting map is changed to driving parameters other than vehicle speed and engine load. By moving the gear accordingly, we were able to obtain a more appropriate gear position. There have been many disclosures regarding this shift map switching (change or correction). For example, Japanese Patent Application Laid-Open No. 62-63251 discloses a technique for correcting a signal from a vehicle speed sensor based on oil temperature. Japanese Patent Publication No. 48-9729 proposes a technology for switching the gear shift map depending on the steering angle. Japanese Patent Application Laid-Open No. 62-37549 discloses a technology for detecting a road surface inclination and changing/correcting a shift map so that the speed change can be performed according to the degree of the road surface inclination. Furthermore, Japanese Patent Laid-Open No. 63-101549 discloses a technique for changing a shift map so as to perform shift control according to the acceleration state of the vehicle. However, in such conventional shift control,
Taking into account various driving parameters (in the case of the above technology, oil temperature, steering angle, road surface inclination, or vehicle acceleration), the numerical values of each driving parameter are monitored in order to obtain the gear position truly requested by the driver. There was a problem in that each time the vehicle was moved, the shift map had to be changed (corrected) depending on the throttle opening and vehicle speed, or a multidimensional map with a huge storage capacity had to be prepared. Since it is virtually impossible to take into account the relationship between each driving parameter in detail, as a result of many corrections, the resulting shift map (=shift stage determined by the shift map) does not necessarily reflect the driver's wishes. Specifically, as mentioned above, when the oil temperature is low, the shift point of the automatic transmission is moved to the high speed side (the shift point is moved to the high speed side). The shifted shift map is selected), and the upshift is performed only when the engine is in a high rotation state.A similar correction is also made to ensure power performance even when the road surface is steeply sloped. ．Therefore, if the oil temperature is low and the road surface is steep, these corrections will be combined,
There will be a situation where the gear will not shift unless the engine is running at a fairly high speed. Further, for example, control is provided to prohibit downshifting to prevent an unexpected increase in driving force when the steering angle is greater than a predetermined value, and control to downshift to increase driving force when the road surface slope is steep. When such controls overlap, problems arise as to how to deal with them. For example, if the accelerator is depressed at a predetermined value or more, if the vehicle acceleration is less than a predetermined value and the vehicle weight is more than a predetermined value, a downshift is performed. No matter how heavy the vehicle acceleration is, if the vehicle acceleration is slightly lower than the given value, downshifting will not be executed, and the shift will be out of touch with the driver's actual needs. Such mutual interference between driving parameters cannot be ignored as the correction control using driving parameters becomes more diverse and complex, but in reality it is extremely difficult to eliminate all interference smoothly. In recent years, in order to improve the sophistication of shift control in automatic transmissions, shift map switching and correction control based on various driving parameters have been increasingly incorporated. However, it has been found that the more such correction control is incorporated, the more the actual shift control may deviate from the shift control actually desired by the driver, which is an ironic result. This is how it came to be. To solve this problem, Japanese Patent Laid-Open No. 63-246546 proposed a technology for determining the gear position by performing fuzzy inference from signals from various detection means (including technology for determining engagement and release of lock-up clutches). It has been done.

[The problem that the invention tries to solve! ! ! ]

However, in general, when determining the gear position by fuzzy inference, the number of driving parameters that are input is large, and as a result, the probability of occurrence of failure in the sensor system, for example, tends to increase accordingly. There was also the problem that a large shift shock occurred, which could cause discomfort to the driver. The present invention has been made in view of such conventional problems, and an object of the present invention is to provide a gear position determination device for an automatic transmission that takes into account that inappropriate gear changes will not occur even in the event of a failure. Suppose that

[Means to solve the problem]

As summarized in FIG. 1, the present invention provides gear position determination for an automatic transmission in which a target gear position is determined by performing fuzzy inference based on signals from sensors of various driving parameters. In the device, means for determining whether or not at least one of the signal systems of the various driving parameters has failed; and when it is determined that the signal system has failed, at least the signal system that has failed; The above object has been achieved by providing means for excluding the target gear position, and means for determining the target gear position by fuzzy reasoning or a gear shift map using only the remaining signal system. In addition, in the present invention, by considering the term "gear stage" in a broad sense, the lock-up clutch is turned on.
, it can also be applied to the case of determining the eye condition of OFF (attachment and release).

[Effect]

As mentioned above, when determining the gear position by fuzzy inference, a large number of driving parameters are often used as input signals in order to make the most of its advantages. There is a tendency for the proportion of people who do this to increase. The present invention includes means for determining whether or not at least one signal system has failed, and when it is determined that the signal system has failed, at least this signal system is excluded before determining the target gear position. That's what I do. For determining failure, various conventionally known methods can be used, as described below. The present invention does not particularly limit the method of determining fail. To exclude a failed signal system, for example, the membership function for that signal system may be rewritten to 1 or O, so that the signal system is essentially not considered. Specifically, if the degree to which a certain condition is satisfied for the signal system is considered by an AND combination, it can be replaced with 1, 0, and if it is taken into consideration by a mooring, it can be replaced with O. On the other hand, if a failure is determined in a particular signal system, the fuzzy inference may be stopped and the target gear position may be determined according to the conventional gear shift map based only on vehicle speed and throttle opening. Note that the present invention does not prohibit configuring each signal system so that a fail itself does not occur. For example, for signals related to engine load and vehicle speed, which are particularly important among signal systems, it would be beneficial to have two or more signal systems and a water system that can back up each other.

【Example】

Figure 2 shows an overall outline of a vehicle automatic transmission to which this embodiment is applied. This automatic transmission includes a torque converter section 2o and an overdrive mechanism section 40 as its transmission sections.
and an underdrive mechanism section 60 with three forward stages and one reverse stage. The torque converter section 2o is a well-known one that includes a pump 21, a turbine 22, a stator 23, and a lock-up clutch 24. The overdrive AM section 4o includes a set of planetary gears including a sun gear 43, a ring gear 44, a planetary pinion 42, and a carrier 41, and the rotational state of this planetary gear is controlled by a clutch C.
It is controlled by O, brake Bo, and one-way clutch Fo. The underdrive mechanism section 60 has a common sangya 6.
1, ring gear 62, 63, planetary pinion 64,
65 and carriers 66 and 67, and the rotation state of the two sets of planetary gear devices and the connection state with the overdrive mechanism are controlled by the clutch CI,
C2, brake 81 to BS, and one-way clutch Fl,
It is controlled by F2. Since the mechanical structure of this transmission section and the hydraulic control system 1.86 that controls it is well known, a detailed explanation will be omitted, with only a skeleton diagram shown in FIG. This automatic transmission includes a transmission section as described above, and a computer (ECU).
Equipped with 84. The computer 84 includes a throttle sensor 80 that detects the throttle opening degree θ to reflect the output (torque) of the engine 1, a vehicle speed sensor (rotational speed sensor of the output dregs 70) 82 that detects the vehicle speed no. a shift range sensor 91 for detecting the operation position of the shift lever (positions such as D, 2, L, R, P, N, etc.); an engine rotation speed sensor 92 for detecting the rotation speed of the engine;
Signals from a steering angle sensor 94 and the like that detect the steering angle of the steering wheel are input. The computer 84 stores in advance a shift map of vehicle speed and throttle opening, determines a target gear according to the control flow described later, and controls the hydraulic control circuit 86 so that the automatic transmission reaches this target gear. Solenoid valve 31, S2 (
SL (for lock-up clutch) and SL (for lock-up clutch) are driven and controlled, and the combination of engagement of each clutch, brake, etc. as shown in Fig. 3 is performed to execute gear changes.
In Fig. 3, the O mark means that the relevant clutch or brake is in the engaged state, and ◎ means that the relevant one-way clutch is activated (state B where power is transmitted from the engine side to the wheel side). This means that it is in a locked state. FIG. 4 shows the control flow when determining the target gear in the computer 84. Here, we will explain the case where the shift lever is selected to the "D" range, which has four forward gears. First, in step S1, the current gear position (current gear position) N, accelerator opening degree θ is determined from a preset gear change map.
A reference gear NX is determined based on cLc and vehicle speed V. The shift map is set in advance using the vehicle speed V and the accelerator opening θαC as parameters, and is determined for each of the shift ranges rl)J and “2” having a plurality of gears. Figure 28 is an example of a shift pattern in the D range, showing vehicle speed V and accelerator opening θ.
The gear shift line is set in a stepped manner in the orthogonal coordinates with the CLC, and the solid line is the upshift shift line.
The broken line is the downshift shift line. Also, 1 in the diagram
, 2, 3, and O/D represent the first gear, second gear, third gear, and overdrive gear (fourth gear), respectively. From this shift map, the current gear N and accelerator opening θQ are determined.
A plurality of judgment vehicle speeds Vl, V2, and V3 are set based on C, and a reference gear N'1g representing the gear to be selected by comparing these judgment vehicle speeds and the actual vehicle speed is set as in the conventional case. It is determined by In other words, when determining this reference gear stage, information on vehicle speed and accelerator opening is naturally required, but if these signal systems are equipped with a backup system, that backup system can be used as in the past. system is used accordingly. The determined vehicle speeds V, , v2, and v3 shown in FIG. 28 are when the current gear N is the third gear and the accelerator opening θaC is 40%; 1'' is determined as the reference gear NX, and if ■+<V≦v2, the 25th
The 11th gear "2" is determined as the reference gear NX, and V2<
If V≦v3, the third gear “3” is determined as the reference gear N-Kyo, and if V s < V, the overdrive gear “4” is determined as the reference gear N-East. The current gear N is read based on the output signals to the solenoid valves S1 and S2.In step SIA, it is determined whether each input signal system has failed. Various well-known methods can be adopted.For example, if the rate of change of the signal over time is monitored and the signal instantaneously becomes 0, it can be determined that a failure has occurred.Also, it is possible to determine that a particular state has failed. It is also possible to determine a failure even if it continues for a long time, which would be impossible to occur.Furthermore, if multiple sensor systems are installed for related targets, for example, the opening degree of the accelerator pedal, the opening of the throttle valve, etc. If equipped with sensors related to opening, intake air pressure, or fuel injection amount,
Since these values have a strong correlation with each other, failure can be determined by comparing the detected values. If a failure occurs in any signal system in this way, the process proceeds to step SIB, where the target gear position k is determined to be the reference gear position N East, and the fuzzy reasoning from step S2 onwards is bypassed. As a result, the target gear position is determined based on the reference gear position Nz obtained from the speed change map of the vehicle speed and engine load as before. In this method, if a failure is detected in even one of the signal systems, all fuzzy inference is stopped and the membership function related to the failure is rewritten in order to determine the target gear using only the gear shift map. This has the advantage that the control flow is extremely simplified since it is not necessary. In addition, in this embodiment, the signal system necessary to determine the reference gear from the shift map, that is, the engine load (throttle opening, intake pressure, etc.)
．． ) and vehicle speed signal system, including the backup system, will not function effectively. However, as mentioned above, when determining the gear ratio using fuzzy calculations, a large number of driving parameters are often used as input signals in order to make the most of the advantages; This becomes effective when the signal system for driving parameters other than engine load and vehicle speed fails. In addition, when the gear position was conventionally determined using a shift map, two systems were installed as vehicle speed sensors, in addition to the system that detects vehicle speed by generating electrical pulses, the vehicle speed (cable) signal from the speedometer was used as a backup system. It was ready for use. However, when using vehicle speed information through fuzzy calculations, an easy-to-process electrical vehicle speed sensor is required, for example to measure vehicle acceleration.
Therefore, it is difficult to perform backups as before. In the present invention, since the reference gear is first determined as before using the shift map, the backup system of the vehicle speed sensor system from the speedometer can be prepared as before when determining the reference gear. It is advantageous. For this purpose, when determining failure in step SIA, the vehicle speed and engine load signal systems are also included in the determination. As a result, even if, for example, the vehicle speed sensor fails and the backup vehicle speed sensor system functions to determine the standard gear position N-kyu, fuzzy inference will be executed using the information from the vehicle speed sensor that failed. This can be prevented. If it is determined in step SIA that no failure has occurred in any signal system, the process proceeds to step S2 and subsequent steps, where fuzzy inference is developed. In step S2, the satisfaction level γQt(j) to be selected for each gear position (j=1, 2, 3J4) is set according to a predetermined fuzzification rule Q1 based on the reference gear position N8. Ru. Fuzzing rule Q
1, the satisfaction level γQ1 (j) is determined based on whether or not it is close to the reference gear N East. For example, for the reference gear N East, the satisfaction level γQl (j)=1
, satisfaction level r Q 1 (j) = O for the gears NX±1 adjacent to the reference gear N8. 5, the satisfaction level γQl (j ＞=
0.25, satisfaction level γQl for gear NX±3
(j) is set to 0.15. FIG. 5 is a diagram showing the degree of satisfaction γQ<(j) of each gear when, for example, the reference gear N8 is the third gear. In the figure, j=1, 2, 3, and 4 are respectively the first far stage, the second gear stage, and
3rd gear, O/D gear (overdrive gear: 4th gear)
Corresponds to Then, in step S3, 'j=1j
After that, in step S4, the number of gears to change ΔN is calculated by subtracting the current gear N from j, and in step S5, each gear is selected according to the actual driving condition using a control rule based on fuzzy inference. The power level of satisfaction γR (j) is calculated. This Irl rule is
It is determined according to the number of change gears ΔN with respect to the current speed change PiN, and subrules A, BB', C, D, E, F, G.
The following four control rules R1 to R4 are set using H. Note that the control rule R1 defines the conditions to be met when ΔN=0, that is, the current gear N is maintained, and the control rule R2 defines the conditions that should be met when ΔN=0, that is, the current gear N is upshifted by one gear. The control rule R3 defines the conditions to be satisfied in the case where ΔN=+2, +3,
That is, it defines the conditions that must be met when upshifting from the current gear N to 2 or 3 gears, and the control rule R4 is ΔN=-1, -2, -3, that is, upshifting from the current gear N to 1
It stipulates the conditions that must be met when downshifting a gear, two gears, or three gears. R1=Aand Band C R2=Aand B' and Cand ((D
and E) or (Fand G) I R 3 = A and B” and Can Fan
d GR4=Aand B' and Cand (
D or H) Also, each of the above sub-rules A, B. B',C
, D, E- F, G, and H each have the following contents. <Subrule A>"Target vehicle drive torque TOxl can be output" This subrule states that the drive torque that can be output at each gear is equal to the engine. Since it is determined by the characteristics, it is determined whether the target vehicle drive torque To East at this time is included in the range of drive torque that can be outputted, and the membership function f^ that represents the degree of satisfaction that satisfies this rule. An example of (To”) is shown in Figure 6. The values C1 and C2 in FIG. 6 are calculated or experimentally determined for each gear, and are set according to the value of "j" corresponding to the gear.The membership function f^(To ”), that is, the satisfaction level is 0.
It is expressed as a number less than or equal to 1, and 1 means that the conditions are completely satisfied. The same applies to each membership function below. Further, the target vehicle drive torque To East is determined by the vehicle speed V and the accelerator opening degree θ, as shown in FIG. 12, for example. It is obtained from a data map etc. with C as a parameter. <Sub-rule B>"Expected rotational speed Ne' is approximately close to target rotational speed Ne" This sub-rule is based on, for example, the target vehicle drive torque TO
The optimum gear position is selected based on the target rotational distance Ne″X when the drive torque To East is relatively small and the drive torque To East can be output at multiple gear positions from 1st gear to O/D gear. Therefore, it is determined whether or not the current target rotational speed Ne is included within the rotational speed range determined for each gear stage with the expected rotational speed Ne' as the center, and this rule is satisfied. Membership function fB representing satisfaction level
An example of (Ne'') is shown by a solid line in FIG. 7. The expected rotational speed Ne' is expressed by a function such as the vehicle speed V and the gear ratio of each gear, and It is set according to the above target rotational speed N eX
For example, as shown in FIG. 13, the target horsepower PS (target vehicle drive torque TO''
It is obtained from a data map etc. with the parameter being (proportional to). Sub-rule B') "Expected rotational speed Ne' is close to target rotational speed N eX" This sub-rule is almost the same as the above sub-rule B, but
Since it is used when changing from the current gear to the next gear, the criteria for this is made stricter, and an example of the membership function f (Ne yi) representing the degree of satisfaction that satisfies this sub-rule is given in the seventh section above. It is shown in the figure with a dashed line. <Subrule C>"Expected engine speed Ne' is within a predetermined allowable range" This subrule is used because if the engine speed Ne is too low, it will induce an engine stall, and if it is too high, it will cause an overrun. The purpose is to prevent the rotation speed from reaching a speed that would impede the operation of the engine, and the membership function f represents the degree of satisfaction with satisfying this sub-rule.
An example of c (Ne') is shown in Figure 8. Values C3 and C4 in FIG. 8 are predetermined according to the characteristics of the installed engine. Sub-rule D〉 “The accelerator is in a steady state” This sub-rule is based on the rate of change #cLC (=dθ.c/
(H), and an example of the membership number fo-(#ac) representing the degree of satisfaction that satisfies this sub-rule is shown by the solid line in Figure 9. show. Sub-rule E ->"The elapsed time T since the last shift is long" This sub-rule is intended to prevent busy shifts in which gears are changed frequently. An example of the number f E (T) is the 10th
It is shown in the figure. Sub-rule F〉 “Accelerator return speed is fast” This sub-rule is based on accelerator opening θ. 0 change rate θ.
This is to determine whether C is negative and relatively large.
An example of the membership function f p (#ac) representing the degree of satisfaction that satisfies this rule is shown in FIG. 9 by the dashed line. <Sable. -Le G>"Not a curve" This sub-rule is to prevent an upshift from occurring when the accelerator is released during a curve, and if the steering angle θS is small, it is determined that the curve is not a curve. FIG. 11 shows an example of the membership function fG《θS) representing the satisfaction level of satisfying this sub-rule. <Sub-rule H>"Accelerator depression speed is fast" This rule is for determining whether the rate of change DαC of accelerator opening θaC is positive and relatively large, and indicates the degree of satisfaction that satisfies this rule. membership function f
An example of H (#ac) is shown in FIG. 9 by the two-dot chain line. In the fuzzy inference method, ``rand'' is defined as algebraic product or minimum operation, etc., and ``0'' is defined as logical sum or maximum operation, etc., but here, if they are defined as algebraic product and maximum operation, respectively, the control The satisfaction level γR(j) of rules R1 to R4 is expressed by the following formula (1).
~(4). γu(j) = f A (Totium)xf B (Ne”)?f
c (Ne') ...<1>
γR (j) = f ^ (T■ East)Xfe
(Ne”) xf c (Ne') xnax
(f o (#ac)xf E (T),
f F (#ac)Xfe(θs))
...《2) γR (j) = f A (To”)Xf a′ (Neii) xf
c (Ne')xf F (#ac)Xfo
(θS〉 ...《3)γR (
j ) =f A (To”) xf e
(Ne east 〉xf c (Ne'
) XflaX (f o (#ac)+
f.
5, the satisfaction level γR at which the second gear should be selected according to the above equation (4) according to the control rule R4 as described above.
(2〉 is found. Then, in this way, the satisfaction level γ
When R(2》) is obtained, in the next step S6,
Based on the satisfaction level γR (2) and the satisfaction level γQl (2) set in step S2, the following formula (5
), the overall satisfaction level γ《2) at which the second gear should be selected is calculated from their algebraic product. γ(2)=γ
R(2)XγQl (2)...(5) j is 1 to 4
If expressed generally as (5A), γ (j) = γR (j) After that, the steps from step S4 onwards are repeated. As a result, the satisfaction levels γ(1), γ(2), γ(3), γ<< 4) are calculated respectively. Specifically, when j=3, ΔN=O, and in step S5, the satisfaction level γR(3) at which the third gear should be selected is determined by the above equation (1) according to the control rule R1. , in the next step S6, the third
The overall satisfaction level γ(3) at which the gear should be selected is given by the above (
5) Calculated according to the formula. Furthermore, in the case of j=4, Δ
N=+1, and in the step S5, the satisfaction level γR(4) at which the O/D stage should be selected is determined by the equation (2) above according to the control rule R2, and then in the next step S6.
The overall satisfaction level γ(
4) is calculated according to equation (5) above. FIG. 14 shows the above step S when the current gear N is 3, that is, the third gear.
FIG. 15 is a diagram showing an example of the satisfaction level γR (j) calculated in step S6, and FIG. 15 is a diagram showing an example of the overall satisfaction level γ(j) calculated in step S6. In the above embodiment, the current gear N is the third gear, so the control rule R3, which defines the conditions to be met when upshifting from the current gear by two gears, is not used. When the gear N is the first gear or the second gear, the third gear
Control rule R3 is used when determining the degree of satisfaction at which gear or O/D gear should be selected. In this way, steps 84 to S8 are repeated, and j=
4, the determination in step S7 becomes NO, and step S9 is subsequently executed. In this step S9, the highest satisfaction level γ(k) of the satisfaction levels γ(j>) of each gear stage calculated in the step S6 is selected. That is, for example, as shown in FIG. When the satisfaction level γ(j) is obtained, γ(2> is selected as γ(k), and in the next step SIO, the target gear to be selected is "k no," of γ(k). That is, in the above example, the second gear stage is determined. Then, in accordance with this determination, the solenoid valves S1 and S2 are respectively energized.
The automatic transmission is downshifted from third gear to second gear. In this way, in the shift control device l4 of this embodiment,
When no failure occurs in any signal system, the target gear is determined by fuzzy inference, but if a failure occurs somewhere, the reference gear N East generated by the shift map becomes the target gear. Since the method is implemented in stages, it is possible to implement fail countermeasures using only an extremely simple flow. Further, in this embodiment, in steps S1 and S2, the satisfaction level γQ1(j) to be selected for each gear stage is set based on a gear shift map with predetermined accelerator opening degree θCLC and vehicle speed V as parameters. At the same time, in step S5, the degree of satisfaction γR (j) to be selected for each gear according to the actual driving condition is calculated according to predetermined control rules R1 to R4, and these degrees of satisfaction γ
The overall satisfaction level γ(j) is calculated in step S6 based on Q1(j) and γR(j). Then, the gear position with the highest overall satisfaction level γ(j) should be selected. Steps S9 and S are selected as the target gear position.
10, the IjL-appropriate shift 111tl is performed in accordance with various driving conditions that cannot be handled only by a shift map using the accelerator opening θαC and vehicle speed V as parameters in normal times when failures do not occur. In addition, since the shift map and fuzzy inference are used together to select the gear, the program is more efficient than the case where multiple correction maps and shift patterns are used in consideration of various driving conditions. Only a small amount is required. This is because when fuzzy inference is used, the amount of programs only increases in proportion to the number of driving parameters to be considered when selecting a gear, whereas when a correction map etc. This is because dividing the state into cases requires a number of maps that are approximately proportional to the power of the number of driving parameters, and the difference becomes more pronounced as the number of driving parameters that must be taken into account increases in order to perform better gear shift control. becomes. In addition, since the basic shift map is predetermined, the amount of programming related to this point is smaller than when using fuzzy inference, and compared to when the shift decision is made based only on fuzzy inference,
The overall program size is reduced accordingly. Incidentally, when the difference in the amount of programs is shown in Fig. 16, when the shift map and fuzzy inference are used together as in this embodiment, the solid line in Fig. 16 shows that the shift map and fuzzy inference are used together, as in this example, and the shift map and fuzzy inference are used together as shown in the solid line in Fig. 16. When the control is performed, the result will be as shown by the dashed line in the same figure. Furthermore, the two-dot chain line in the same figure shows the case where shift control is performed only using the shift map and correction map. Furthermore, in this embodiment, since the gear position with the highest overall satisfaction level γ(j) calculated for each gear position is selected, the shift map is created based on the calculation result by fuzzy inference, for example. Unlike in the case of correction, there is no need to perform single point conversion (defuzzification) using the centroid method or the like in order to obtain a specific correction amount from the calculation results, and the amount of programming can be reduced in this respect as well. However, when determining the gear, it is also possible to find the center of gravity, center of area, etc. of the satisfaction level γ(j) using the center of gravity method, area method, etc., and select the gear that is closest to it. Furthermore, in this embodiment, the membership functions in fuzzy inference are defined with slopes as shown in FIGS. 6 to 11, so that
By appropriately determining the slope, it becomes possible to perform gear change control that is more in line with the driver's senses. Furthermore, according to such shift control based on fuzzy reasoning, even when including a parameter that is generally difficult to measure with high precision, such as the speed of change in accelerator opening bcLc, the shift control can be carried out favorably. You can get the following advantages. Next, a second embodiment of the present invention will be described using FIG. 17. As is clear from FIGS. 1-7, this second embodiment provides step SS2 in place of step S2 of the first embodiment {FIG. 4}, and provides step SS5 in place of steps S5 and S6. Since this is a new version, only the changed parts will be explained. First, in step SS2, the degree of satisfaction γQ+(j>, γoz(
j>, γa3(j) is set. Fuzzing rule Q
1 is the reference gear stage N as in the first embodiment (Fig. 4).
The degree of satisfaction is determined based on whether or not it is close to the reference gear position NX.The fuzzification rule Q2 determines the degree of satisfaction based on whether or not it is approximately close to the reference gear position The degree of satisfaction is determined based on whether or not it is very close to stage NX. Reference gear stage N.
Satisfaction levels γoz(j) and γQa for each gear set by fuzzing rules Q2 and Q3 when x is the third gear
Examples of (j> are shown in FIGS. 18 and 19, respectively (γ
For Qt(j), see Figure 5). Also, step SS5 is based on the fuzzing rules Q1 and Q
2. Using five control rules RI to RV based on fuzzy inference incorporating Q3, the overall satisfaction level γ(j) at which each gear should be selected according to the actual driving condition is calculated. These control rules RI to RV are determined according to the number of change gears ΔN with respect to the current gear N, and include the fuzzing rules Ql, Q2, Q3 and the sub-rules D, E,
The following settings are made using F, G, H and a new sub-rule ■. That is, the above sub-rule I is "the accelerator is fully closed"
An example of the satisfaction level lx(θ(IQ)) that satisfies the sub-runo is shown in FIG. The iIFI control rule Rff defines the condition that must be satisfied when ΔN=-1, that is, when upshifting by one gear from the current gear N. RI is ΔN=+2, +3, that is, when upshifting from the current gear FN to 2 or 3 gears (conditions that must be met are defined), and the control run IRIV is ΔN=-1, that is, when shifting up 2 or 3 gears from the current gear N. The control rule RV specifies the conditions to be met when downshifting, and the control rule RV is ΔN=-2, 12, that is, 2 from the current gear N.
This defines the conditions that must be met when downshifting a gear or three gears.
FIRI[=Q3andF RIV=Q1and (H or I)RV=Q3
and Gand H. Also, if we define "rand" as an algebraic product and rorJ as a maximum operation, then the satisfaction level γ(
j ) are calculated according to the following equations (6) to (1o), respectively. γ (j) = γoz (j) ... (6) γ (
j > = γo + (j) XllaX (f o (
/}ac)xf E (T). f F (#cc
))...(7) r (j)=793 (j)xf F (#crc
)...(8) γ(j> = γQ t (j) xmax (f H (#
cc). f Seal (θ cL c))
...《9》γ(j) =ros(j)xf O(θS) xf H(#ac) − (10) Here, we will specifically explain the case where the current gear stage N is “3”. , when j=1, ΔN=-2, and the control rule RV
Accordingly, the overall satisfaction level γ(1) at which the l-th gear should be selected is determined by the above equation (1o), and when j=2, ΔN=-1, and according to the control rule RIV, by the above equation 9, The overall satisfaction level γ(2) at which the second gear should be selected is determined, and when j=3, ΔN=O, and the third gear is selected according to the above equation (6) according to the control rule RI. The overall satisfaction level γ(3) is calculated,
When j=4, ΔN=+1, and the control rule RI [
Accordingly, the overall satisfaction level γ(4) at which the O/D stage should be selected can be determined using equation (7) above. Then, among these satisfaction degrees γ(j), the gear position k with the highest degree of satisfaction is selected in step S10, SIA,
Regarding the purpose of SIB (1ml capacity and effect), please refer to Part 1 above.
This is exactly the same as the example. In this third embodiment, as is clear from the above-mentioned control rule R, if the reference gear N true found from the shift map in normal times is the same as the current gear N, then the current gear N The satisfaction level γ(j) for which should be selected is 1, and the current gear N is maintained regardless of other driving conditions. Further, according to the shift control device of the third embodiment, in addition to obtaining the same effects as the first embodiment in steps SIA and SIB, three types of satisfaction levels γQ1(j) can be obtained in step S82. , γQ2(j), and γas<j), it is possible to perform gear change control that is more in line with the driver's wishes, and the satisfaction level γa1(j
), γQ2 (j), rQs (j) are incorporated into the control rule R~Rv to calculate the overall satisfaction level γ(j)
Since it is designed to calculate at once, the amount of program can be further reduced. By the way, the present invention can also be applied to the control of engagement and disengagement of the lock-up clutch 24 (FIG. 2) by considering "gear stage" in a broad sense. This example is the 23rd
Explain using a diagram. First, in step S1, the reference command value Lu East is set to the lock-up clutch shift map (
The map will be read according to the map. The shift map of the lock-up clutch 24 is set as shown in FIG. 29, and is determined in the same manner as in the past. In step SIA, it is determined whether a fail has occurred in various signal systems. If a failure occurs, the flow proceeds to SIB, all flows related to fuzzy reasoning are bypassed, and the engagement state of the lock-up clutch 24 is controlled based on the reference command value of the shift map. on the other hand,
When it is determined that no failure has occurred in any signal system, fuzzy inference from step S2 onwards is developed. The basic way of developing this fuzzy inference is explained in the fourth section.
The control flow is exactly the same as the one shown in the figure. Of course, the specific contents of each control rule and sub-rule will be different, but since the specific contents themselves are outside the scope of the present invention, a detailed explanation will be omitted. To briefly explain the control flow, first, γQ (ON) and γo (OFF) are set based on the reference command value Lu.
is set, and in step S3, γR (ON) and γR (OFF) are determined by fuzzy inference from various driving conditions.
is calculated. In step S4, γp (ON
) and γQ(ON) to obtain the overall satisfaction level γ(ON) when the lock-up clutch 24 is turned on (engaged), and similarly calculate γR(OF
By multiplying F) by γca (OFF), calculate the overall satisfaction level γ (OFF) when the lock-up clutch is turned OFF (released). In step S5, both γ(ON) and γ(OFF) are compared, and the larger one is set as γ(k), and in step S6, the lock-up clutch 24 is controlled to the state of k by the solenoid valve sL. FIG. 24 shows an example in which a development basically similar to the fuzzy reasoning shown in FIG. 17 is applied to the control of the lock-up clutch 24. Since the explanation will be redundant, only a flowchart is shown in FIG. 24, and detailed explanation will be omitted. 23 and 24, it is possible to obtain the same effect against failures as explained in FIGS. 4 and 17 regarding the engagement and release control of the lock-up clutch 24. All of the above embodiments are shift maps. It was assumed that the standard gear position N (or standard command value Lu) obtained by It can be applied to minimize the effect even when the accelerator opening degree θ.C or vehicle speed ■ fails including the backup system.For example, when it is determined that the accelerator opening degree θαC has failed, the member ship By rewriting the function from the state shown in Figure 9 to the state shown in Figure 25, for example, it becomes possible to perform fuzzy inference after excluding information regarding the accelerator opening. Specifically, do you rewrite everything to "1"?
Or, you can set them all to 0, but which one you choose depends on how this membership function is reflected in the control rules. In general, if it is singly bonded with "0", O is preferable, otherwise rlJ is preferable. As yet another embodiment of the present invention, there is a method of incorporating the fail judgment itself into the control rules. For example, if a failure of the vehicle speed V is determined based on a sudden change in the vehicle speed V, then the membership function can be calculated as shown in Fig. 26, where = dV/dt, that is, the vehicle speed difference between the dt time. fv(
l1) (vehicle speed fail) and fw (I1) (vehicles fail), and in the case of fuzzy inference as shown in Fig. 4, Equations 1 to (4), In the case of fuzzy inference as shown, rules can be combined for equations (6) to (10) as shown in FIG. 27 (A) or (B), respectively. Then, γR (j) in step S5 of FIG. 4 is expressed as nax(
rp (j) x fv (1?I 冫, fw〈11〉}, while SS5 in Fig.
γ(j ＞, nax(r (j ＞ x tv <
+1>, fw(l1)×γQ1 (j)). As a result, now, for example, in FIG. 26, the value of 11 is a1
When it is, the satisfaction level that the vehicle speed is not a failure is fv
(11) is 1, and the satisfaction level fw (11
> is 0, so “
The control flow that is virtually the same as the flow in the case of "No Fail" is executed. On the other hand, for example, when 11 is the value of a2, the value of the satisfaction level fv (l1) that the vehicle speed is not a failure is equal to the value of the satisfaction degree fw (11) that the OJ vehicle speed is a failure, so the value is 1. , reference gear N
γR (j
) will be set. Therefore, the overall γ(j
) is controlled so as to match the standard variable noodle stage NX. In other words, a control flow that is substantially the same as the flow in the "fail" scenario shown in FIG. 4 or FIG. In this case, the result of intermediate fuzzy inference is adopted. It is clear that the same method can be adopted for other signal system failures. Although the present invention has been described in detail based on
, Q2, and Q3 have been explained, but in order to obtain the satisfaction levels γQ4 and γo5 shown in FIGS. 21 and 22, the gear ratio should be It is also possible to adopt various other fuzzification rules, such as a fuzzification rule that emphasizes the small high gear side and, conversely, a fuzzification rule that emphasizes the low gear side.Also, in the above embodiment, the reference speed change The satisfaction levels of other gears are set with gear NX set to 1, but the judgment vehicle speeds V1, V 2 , and V determined from the gear shift map
Based on 3 and the actual vehicle speed 2, the degree of satisfaction for each gear stage can be set even more precisely using an arithmetic expression or the like. In addition, in the above embodiment, "rand" in fuzzy inference,
Although we have explained the case where rorJ is defined as an algebraic product and a maximum operation, respectively, these definitions and inference methods may be changed as appropriate. Furthermore, in the embodiments described above, four or five control rules Rl to R4 and RI to RV were defined, but the number and contents of these IIJil rules, that is, sub-rules, can be changed as appropriate. As described above, the present invention does not specifically limit how fuzzy inference is performed in consideration of various driving conditions. Further, although the shift map in the above embodiment is set in a stepped manner in the orthogonal coordinates of the accelerator opening θcLc and the vehicle speed V, it is also possible to set a shift map in a straight line, curve, curved line, or the like. Note that it is of course possible to calculate the reference gear Nm after correcting this gear shift map according to engine specifications, driver preference (manual setting), etc.

[Brief Description of the Drawings] Fig. 1 is a block diagram showing the gist of the present invention, and Fig. 2 is a block diagram showing the gist of the present invention.
FIG. 3 is an overall schematic diagram showing the configuration of an automatic transmission equipped with a speed change control device that is an embodiment of the present invention. 4 is a flowchart explaining the operation of the automatic transmission of FIG. 1; FIG. 5 is a diagram showing the engagement state of the coupling elements; FIG. 6 to 11 are diagrams showing an example of the membership function of the control rule used in step S5 of FIG. 3, respectively. FIG. FIG. 13 is a diagram showing an example of a data map for determining the vehicle drive torque. FIG. 14 is a diagram showing an example of the data map for determining the target rotation speed of the engine. FIG. 14 is a diagram showing an example of the data map for determining the target rotational speed of the engine. FIG. 15 is a diagram showing an example of the inference result in step S6 of FIG. 3. FIG. 16 is a diagram showing an example of the inference result in step S6 of FIG. Figure 17 is a flowchart explaining the operation of the second embodiment of the present invention; Figures 18 and 19 are diagrams showing lines that are different from each other in step SS2 of Figure 17; 20 is a diagram showing an example of each gear stage set according to the fuzzification rule; FIG. 20 is a diagram showing an example of the membership function of the control rule used in step SS5 of FIG. 17; FIGS. 21 and 22 are diagrams showing an example of the degree of satisfaction of each gear stage set according to the fuzzing rules, which are different from those in FIGS. 5, 18, and 19, respectively. A flowchart explaining the operation when the invention is applied to the control of a lock-up clutch, Fig. 25 1 is a diagram showing the rewriting of the membership function when the signal system related to the accelerator opening fails, Fig. 26 is , A diagram showing the member function 1 when incorporating the judgment when the vehicle speed signal system fails into fuzzy inference. Figures 27 (A) and (B) respectively take into consideration the case when the vehicle speed signal system fails. A schematic diagram expressing the control rules, Fig. 28 is a diagram showing an example of a shift map of vehicle speed - throttle opening (engine load), and Fig. 29 is an example of a shift map (engagement map) of a lock-up clutch. This is a line diagram showing . 14... Speed change control device, 32... Microcomputer, γQ+(j). γa2(j), γas(j)... Satisfaction level of each gear stage set by the setting means, γR(j
)... Satisfaction level of each gear position determined by the first calculation means, γ(j)... Overall satisfaction level of each gear level, f^, f
day, f day', f c, f o. f E, f F, f
o, fH, f,...Membership function, NX...
Reference gear position, k...Target gear position.

Claims (1)

[Claims] (1) In a gear position determination device for an automatic transmission configured to determine a target gear position by performing fuzzy inference based on signals from sensors for various driving parameters, at least one of the signal systems for the various driving parameters; Regarding one, means for determining whether or not the relevant signal system has failed, a means for excluding at least the failed signal system when it is determined that the relevant signal system has failed, and a means for using the remaining signal system. A gear position determination device for an automatic transmission, comprising: means for determining the target gear position by fuzzy inference or a gear change map.