JP2010252573A - Method of detecting temperature of inverter, control method, and control unit of variable magnetic flux motor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain safety and reliability in terms of operation of inverters by further fully analyzing transient elements to temperature of semiconductor power elements in the inverters, and by precisely estimating a change in temperature at junction sections of the semiconductor power elements. <P>SOLUTION: A modeling circuit, which includes a thermal resistance and a thermal capacity between a junction section of the semiconductor power element and an on-chip sensor for temperature measurement, is defined, an output of the modeling circuit responding to a direct output of the on-chip sensor is set as an estimated loss, and the estimated loss is added to the direct output of the on-chip sensor as estimated measurement temperature data. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an inverter temperature detection method and a control device for a variable magnetic flux motor, and improves the temperature detection performance in the inverter to obtain operational reliability and stabilization of control of the variable magnetic flux motor driven by the inverter. It is.

  There is a technique disclosed in Patent Document 1 as a temperature detection device for a joint portion of a semiconductor element. In this technique, a diode for temperature measurement is added to a semiconductor element. The temperature detection circuit detects the measured temperature data of the position of the diode based on the temperature characteristic of the forward voltage drop of the temperature measuring diode. On the other hand, the current detection circuit measures the sense current flowing through the current sensing emitter of the semiconductor element. The calculation unit calculates correction temperature data based on the sense current value and adds the correction temperature data to the measured temperature data to obtain the junction temperature data.

JP 2004-117111 A

  The above technique obtains estimated junction temperature data based on the output terminal voltage derived from the temperature measuring diode and the sense current flowing through the current sensing emitter. Therefore, when the estimated junction temperature data is obtained, it is a result of the overall heat capacity of the semiconductor device package being integrated. In this case, no consideration is given to transient factors such as thermal resistance and thermal capacity between the junctions of the diode and the semiconductor power device to be measured.

  Recently, there has been a demand for downsizing and lowering the price of inverters. In particular, an inverter applied to a hybrid vehicle and an electric vehicle is required to be reduced in size and price. On the other hand, in order to switch the magnetic force of the variable magnetic flux motor, it is necessary for a miniaturized inverter to instantaneously pass a current exceeding the rated value through the inverter in order to obtain power. Here, in a semiconductor power element used for an inverter, if a current exceeding the rating is instantaneously passed in a state where the temperature is abnormally high, the semiconductor power element is destroyed. In order to solve such a problem, it is necessary to accurately measure and monitor the temperature of the semiconductor power element.

  Therefore, the present invention can further analyze the transient element with respect to the temperature of the semiconductor power element of the inverter and estimate the temperature change with high accuracy, and the operational safety and reliability of the inverter and the stabilization of the motor control. It is an object to provide an inverter temperature detection method and a variable magnetic flux motor.

  To solve the above problems, the present invention defines a modeling circuit including a thermal resistance and a heat capacity between a junction of a semiconductor power element and an on-chip sensor for temperature measurement, and responds to the direct output of the on-chip sensor. The output of the modeling circuit is an estimated loss, and the estimated loss is added to the direct output of the on-chip sensor to obtain estimated measurement temperature data.

  According to the above solution, by defining the modeling circuit, the transient element with respect to the temperature of the semiconductor power element of the inverter can be analyzed in more detail, and the temperature change of the junction part of the semiconductor power element can be accurately estimated. Therefore, operational safety and reliability of the inverter can be obtained.

It is circuit explanatory drawing around the to-be-measured semiconductor switch element which employ | adopts the temperature detection method which concerns on one Example of this invention. It is explanatory drawing which shows the example of the modeling circuit which concerns on one Example of this invention. It is composition explanatory drawing which shows a specific example which implemented the temperature detection method which concerns on one Example of this invention. It is composition explanatory drawing which shows a specific example which implemented the temperature detection method which concerns on the other Example of this invention. FIG. 5 is an operation characteristic diagram shown for explaining an operation example of the embodiment of FIG. 4. 1 is a configuration explanatory diagram of a control device for a variable magnetic flux motor according to an embodiment of the present invention. FIG. It is a simple model figure of a variable magnetic flux motor. It is a figure which shows the cross section of the rotor of the variable magnetic flux motor of FIG. It is a figure which shows the BH characteristic of the variable magnetic flux motor of FIG. It is a figure which shows the BH characteristic of various materials. It is a figure which shows the internal structural example of the magnetization request | requirement production | generation part 29 of FIG. It is a figure which shows the internal structural example of the variable magnetic flux control part 13 of FIG. It is a figure which shows the control timing of the variable magnetic flux motor of FIG. It is explanatory drawing of the compound apparatus to which this invention is applied.

  Hereinafter, embodiments of the present invention will be described in detail. FIG. 1 shows an embodiment of the present invention, and shows the periphery of a semiconductor power element of an inverter to which the present invention is applied. The semiconductor power element 111 is an insulated gate bipolar transistor (hereinafter abbreviated as IGBT) and is on / off controlled by a control signal from the gate switch circuit 112. Reference numeral 113 denotes a power supply line. Here, the junction temperature of the semiconductor power element 111 is measured by the temperature sensor 121. The temperature sensor 121 is a PN junction diode and utilizes the fact that the forward voltage drop has temperature characteristics.

  FIG. 1 shows a typical basic configuration, and a temperature sensor is provided for each of a plurality of semiconductor power elements in the inverter.

  The output of the temperature sensor 121 is input to the temperature information processing unit 122. The temperature information processing unit 122 can measure the temperature of the junction of the semiconductor power element 11 with high accuracy. The following points have been devised to increase the accuracy.

  First, due to the arrangement relationship between the temperature sensor 121 and the semiconductor power element 11, a difference occurs in the measured temperature. Therefore, in the present invention, the modeling circuit 30 corresponding to a space (a portion where an insulating material or the like exists) as shown by a dotted line in FIG. 1 is designed. The modeling circuit 130 includes an estimated IGBT loss that becomes a heat loss.

  FIG. 2 shows an image configuration example of the modeling circuit 130. For example, heat resistance and heat capacity are replaced with electric resistance and a capacitor. A modeling circuit 130 is formed by connecting a plurality of parallel circuits of resistors and capacitors in series. An output from the temperature sensor 121 is input to the input unit 130a as junction temperature data. Then, the junction temperature data is corrected by the modeling circuit 130 and output as correction temperature data to the output unit 130b.

  The above modeling circuit 130 is actually an arithmetic circuit, and is incorporated in the temperature information processing unit 122 as shown in FIG. The output from the temperature sensor 121 is input to the modeling circuit 130 and the adder 124 via the terminal 23a. The correction temperature data output from the modeling circuit 130 is input to the adder 124. The output of the adder 124 is led to the output terminal 123b.

  Here, the modeling circuit 130 is designed for resistance and capacitance based on the estimated IGBT loss parameter so that the heat loss is restored and reflected in the output. The heat loss can be roughly classified according to the conduction loss Pon and the switching loss Psw.

Conduction loss: Pon ＝ Ic × Vce
Switching loss: Psw = (Eon + Eoff) × fsw
Total loss: Ploss = Pon + Psw
Ic: Collector current (A),
Vce: Collector-emitter voltage (V)
Eon: Conduction voltage from off to on
Eoff: conduction voltage from on to off fsw: switching frequency The above loss Ploss is calculated by the rated value of the IGBT and the use conditions (frequency, etc.). Therefore, in the modeling circuit 130, coefficients of resistance and capacitance are set in advance so that Ploss is obtained according to the variable elements (voltage, current, frequency) of the above calculation formula.

  As described above, in the present invention, a modeling circuit including a thermal resistance and a heat capacity between a junction part of a semiconductor power element as an element to be measured and a temperature sensor as an on-chip sensor is defined. Then, the output of the modeling circuit responding to the direct output of the temperature sensor is used as an estimated loss (temperature data for correction), and this estimated loss is added to the direct output of the temperature sensor to obtain estimated measured temperature data. .

  FIG. 4 shows an example of use of the temperature information processing unit 122 that incorporates the modeling circuit 130, and shows an example of use in an inverter. The junction temperature data is input to the input terminal 123a. The junction temperature data is input to the proportional constant calculation unit 141 and also input to the subtractor 142. The proportional constant calculation unit 141 corresponds to the total resistance value of the previous modeling circuit 130.

  In the subtractor 142, the junction temperature data input from the junction upper limit temperature data is subtracted and output. That is, the output of the subtractor 142 indicates how close the currently detected temperature is to the allowable temperature.

  The output of the subtractor 142 is input to an integral constant calculator 143 that repeats integration every predetermined time. The integral constant calculation unit 143 corresponds to the total capacity value of the previous modeling circuit 130. A limit is given to the integral value. The integral value is input to the subtracter 44. The subtracter 144 subtracts the output of the integral constant calculator 143 from the output of the proportional constant calculator 141. The output of the subtractor 144 is input to the correction unit 145. The correction unit 145 subtracts the output from the subtractor 144 from the command current that commands the current flowing through the inverter, and outputs a corrected command current.

  That is, the output of the proportionality constant calculation unit 141 works in a direction to suppress the current of the inverter in proportion to the increase in the junction temperature of the semiconductor power element. In other words, it is estimated that the junction temperature of the semiconductor power element is higher than the measurement temperature (estimated systematically by the modeling circuit), so that the current of the semiconductor power element is suppressed.

  However, the current suppression is relaxed by the output (integrated value) of the integral constant calculation unit 143 so as not to be excessively suppressed. That is, in this embodiment, the current suppression is alleviated by subtracting the output (integral value) of the integral constant calculator 143 from the output (proportional value) of the proportional constant calculator 41.

  In this case, the integral constant calculation unit 143 integrates temperature data (output of the subtractor 142) that can be tolerated up to the limit. Therefore, mitigating the current suppression corresponds to using the IGBT up to the junction temperature limit. .

  FIG. 5 is a diagram illustrating the operating characteristics of the temperature information processing unit 122 described above. (A) shows the change in the junction temperature, and (B) shows the output of the proportional constant calculator 141 and the output of the integral constant calculator 143. (C) shows how the command current changes. It is understood that the temperature estimation operation is started as the temperature of the junction increases, and current suppression for the command current is performed accordingly.

  FIG. 6 shows a drive system for a variable magnetic flux motor to which the present invention is applied. First, before explaining the figure, a variable magnetic flux motor as a permanent magnet synchronous motor will be explained.

  FIG. 7 is a model of the variable magnetic flux motor 1 as a simple image. The stator side may be considered the same as a conventional motor. On the rotor 51 side, as permanent magnets, there are a fixed magnet FMG whose magnetic flux density is fixed and a variable magnet VMG whose magnetic flux density is variable. The conventional permanent magnet synchronous motor (PM motor) is only the former fixed magnet FMG, whereas the variable magnetic flux motor 1 in this embodiment is provided with a variable magnet VMG.

  Here, description is added about a fixed magnet and a variable magnet. The permanent magnet means that the amount of magnetic flux is not constant and the magnetic flux density is not substantially changed by a current supplied from an inverter or the like in a state close to normal rated operation. On the other hand, the above-described permanent magnet having a variable magnetic flux density, that is, a variable magnet refers to a magnet whose magnetic flux density changes due to a current that can be passed through an inverter or the like even under the above operating conditions. Such a variable magnet can be designed within a certain range depending on the material and structure of the magnetic material.

  Modern PM motors often use neodymium (NdFeB) magnets with a high residual magnetic flux density Br. In the case of this magnet, since the residual magnetic flux density Br is as high as about 1.2 T, it is possible to output a large torque with a small device size, and it is suitable for a hybrid vehicle HEV or a train that requires a high output and a small motor. . In the case of a conventional PM motor, it is a requirement that it is not demagnetized by a normal current. However, this neodymium magnet (NdFeB) has a very high holding force Hc of about 1000 kA / m. It is an optimal magnetic material. This is because a magnet having a large residual magnetic flux density and a large coercive force is selected for the PM motor.

  On the other hand, a magnetic material such as Alnico AlNiCo (Hc = 60 to 120 kA / m) or FeCrCo magnet (Hc = about 60 kA / m) having a high residual magnetic flux density and a small coercive force Hc is used as a variable magnet. The magnetic flux density (the amount of magnetic flux) of a neodymium magnet is almost constant depending on the normal amount of current (meaning the amount of current that flows when a conventional PM motor is driven by an inverter), but a variable magnet such as an Alnico AlNiCo magnet The magnetic flux density (magnetic flux amount) is variable.

  Strictly speaking, since the neodymium magnet is used in the reversible region, the magnetic flux density fluctuates within a very small range, but returns to the original value when the inverter current disappears. On the other hand, since the variable magnet is used up to the irreversible region, the initial value is not obtained even if the inverter current is lost. The amount of magnetic flux of the alnico magnet, which is the variable magnet VMG in FIG. 7, is almost zero in the Q-axis direction only by the amount in the D-axis direction changing.

  FIG. 8 shows a specific configuration example of the variable magnetic flux motor 1. The rotor (rotor) 51 has a configuration in which a high coercivity permanent magnet 54 such as a neodymium magnet (NdFeB) and a low coercivity permanent magnet 53 such as an alnico magnet (AlNiCo) are combined in a rotor core 52. It is. The low coercive force permanent magnets 53 that are the variable magnets VMG are disposed on both sides of the magnetic pole part 55 of the rotor core 52 in the radial direction in the boundary area with the adjacent magnetic pole part 55. The high coercive force magnet 54 that is the fixed magnet FMG is arranged in a direction perpendicular to the diameter in the magnetic pole portion 55 of the rotor core 52. With this structure, the low coercive force permanent magnet 53, which is the variable magnet VMG, is magnetized by the D-axis current without being affected by the Q-axis current because the Q-axis direction and the magnetization direction thereof are orthogonal to each other.

  FIG. 9 illustrates the BH characteristics (magnetic flux density-magnetization characteristics) of the fixed magnet and the variable magnet. FIG. 10 shows only the second quadrant of FIG. 9 in a quantitatively correct relationship. Furthermore, BH characteristics of various materials are shown.

  In the case of a neodymium magnet and an Alnico magnet, there is no significant difference between their residual magnetic flux densities Br1 and Br2, but for the coercive forces Hc1 and Hc2, Hc1 of the Alnico magnet (AlNiCo) is equal to Hc2 of the neodymium magnet (NdFeB). From 1/15 to 1/8, the Hc1 of the FeCrCo magnet is 1/15.

  In the conventional PM motor drive system, the magnetization region due to the output current of the inverter is sufficiently smaller than the coercive force of the neodymium magnet (NdFeB), and is utilized in the reversible range of its magnetization characteristics. However, since the coercive force of the variable magnet is small as described above, it can be used in the irreversible region (even if the current is zero, it does not return to the magnetic flux density B before the current application) in the inverter output current range. Thus, the magnetic flux density (magnetic flux amount) can be made variable.

An equivalent simple model of the dynamic characteristics of the variable magnetic flux motor 1 is shown in equation (1). The model is a model on the DQ axis rotational coordinate system in which the D axis is given as the magnet magnetic flux direction and the Q axis is perpendicular to the D axis.

  Here, R1: winding resistance, Ld: D-axis inductance, Lq: Q-axis inductance, Φfix: amount of magnetic flux of the fixed magnet, Φvar: amount of magnetic flux of the variable magnet, and ω1: inverter frequency.

  FIG. 6 shows a main circuit and a control circuit of the variable magnetic flux drive system according to the first embodiment. The main circuit includes a DC power supply 3, an inverter 4 that converts DC power into AC power, and a variable magnetic flux motor 1 that is driven by AC power (drive current) of the inverter 4. The main circuit is provided with an AC current detector 2 for detecting motor power and a speed detector 18 for detecting motor speed.

  Specifically, a three-phase drive current from the inverter 4 is supplied to the status side winding of the variable magnetic flux motor 1. For example, changes in the two-phase drive current are detected by the current detection devices 2U and 2W. The current detection devices 2U and 2W have detection circuits using, for example, Hall elements. The detected current is input to the coordinate conversion unit 7. The coordinate conversion unit 7 converts the AC detection currents Iu and Iw of the current detector 2 into two-axis DQ axes, converts them into DQ-axis current detection values Id and Iq, and inputs them to the voltage command calculation unit 10.

  Next, the control circuit will be described. The inputs here are the operation command Run * and the torque command Tm *. The operation command generation unit 16 receives the operation command Run * and the protection signal PROT determined by the protection determination unit 17 and generates and outputs an operation state flag Run. Basically, when the operation command is entered (Run * = 1), the operation state flag Run is set to the operation state (Run = 1), and when the operation command instructs to stop (Run * = 0), The operation state flag Run is set to the stop state (Run = 0). Further, in the case of protection detection (PROT = 1), even if the operation command Run * = 1, the operation state is set to the stop state Run = 0.

  The gate command generation unit 15 receives the operation state flag Run, and generates and outputs a gate command Gst to the switching element included in the inverter 4. In the gate command generation unit 15, when the operation state flag Run changes from stop (Run = 0) to operation (Run = 1), the gate start (Gst = 1) is immediately started, and the operation state flag Run operates (Run = In the case of changing from 1) to stop (Run = 0), the gate is turned off (Gst = 0) after a predetermined time has elapsed.

  The current reference calculation unit 11 outputs command currents Id * and Iq * that determine the change timing and amount of magnetic flux of the coil of the variable magnetic flux motor 1. The command currents Id * and Iq * are output toward the voltage command calculation unit 10.

  Although a temperature information processing unit 122 for detecting the temperature of the semiconductor power element of the inverter 4 is shown, the relationship between the temperature information processing unit 122 and each unit will be described later.

The magnetic flux command calculation unit 12 receives the operation state flag Run and the inverter frequency ω1, that is, the rotor rotational frequency ωR, and generates and outputs a magnetic flux command Φ * as shown in the following equation (2), for example. That is, when the operation is stopped (Run = 0), the magnetic flux command Φ * is set to the minimum Φmin, the operation state (Run = 1), and the rotational frequency ωR is lower than the predetermined value, the magnetic flux The command Φ * is set to the maximum Φmax, and when the speed is higher than a predetermined value, the magnetic flux command Φ * is set to the minimum Φmin.

  Here, Φmin is the minimum amount of magnetic flux (> 0) that can be taken as the variable magnetic flux motor 1, Φmax is the maximum amount of magnetic flux that can be taken as the variable magnetic flux motor 1, and ωA is a predetermined rotational frequency. The setting of the magnetic flux amounts Φmin and Φmax will be described later in the variable magnetic flux controller 13.

The current reference calculation unit 11 receives the torque command Tm * and the magnetic flux command Φ * and calculates the D-axis current reference IdR and the Q-axis current reference IqR as in the following equations (3) and (4).

  The expressions (3) and (4) are arithmetic expressions assuming that the motor reluctance torque is not used and the number of motor poles is zero. Even a salient pole motor having a difference ΔL between the D-axis inductance Ld and the Q-axis inductance Lq may be a non-salient pole type motor having no difference.

However, it is effective to consider the reluctance torque when optimizing the efficiency and considering the maximum output at a predetermined current. In this case, for example, the calculation is performed as follows.

  Here, K is the ratio of the D-axis current and the Q-axis current, and is a value that varies depending on the application, such as the aforementioned efficiency optimization and maximum output. To optimize, it takes a function form and uses torque, speed, etc. as its arguments. In addition, simple approximation or a table can be used. Further, the magnetic flux command Φ * in the equation (5) can be operated even if a magnetic flux estimated value Φh described later is used.

  FIG. 11 shows a detailed configuration of the magnetization request generator 29. The block in FIG. 11 is assumed to be controlled every predetermined time by the control microcomputer. The magnetic flux command Φ * is input to the previous value holding unit 31 and the value is held. The output of the previous value holding unit 31 is the previously stored magnetic flux command Φ *, and is input to the change determination unit 30 together with the current magnetic flux command value Φ *. The change determination unit 30 outputs 1 when there are two input changes, and outputs 0 when there is no change. That is, 1 is set only when the magnetic flux command Φ * changes. A circuit similar to the above is provided for the operation state flag Run instead of the magnetic flux command Φ *. The outputs of the two change determination units 30 and 34 are input to the logical sum operation unit (OR) 32, and the logical sum of these is output as the magnetization request flag FCreq.

  The magnetization request flag FCreq, which is the output of the magnetization request generation unit 29, becomes a magnetization request (FCreq = 1) when the magnetic flux command Φ * changes or when the operation state flag Run changes, otherwise there is no request ( FCreq = 0). The state in which the operation state flag Run changes is when the inverter starts, stops, or stops due to protection. Further, although the magnetic flux command Φ * is used here, the magnetization request FCreq may be generated by a change in a magnetization current command Im * (output of the magnetization current table 27) of the variable magnetic flux control unit 13 described later.

  FIG. 12 shows a detailed configuration of the variable magnetic flux control unit 13. The variable magnetic flux control unit 13 receives the magnetic flux command Φ * that is the output of the magnetic flux command calculation unit 12 and outputs a D-axis magnetization current difference amount ΔIdm * that corrects the D-axis current reference IdR. The magnetization current difference amount ΔIdm * is input to the adder 14 to correct the reference current from the current reference calculation unit 11.

  The generation of the magnetizing current difference amount ΔIdm * is performed by the following arithmetic processing. In order to magnetize the variable magnet VMG, a predetermined magnetization current command Im * may be obtained in accordance with the BH characteristics of the variable magnet shown in FIG. In particular, the magnitude of the magnetization current command Im * is set to be equal to or greater than H1sat in FIG. 9, that is, the magnetization saturation region of the variable magnet.

Since the magnetization current flows to the magnetization saturation region, the magnetic flux amount Φmin or Φmax to be set by the magnetic flux command calculation unit 12 is set to the maximum value (saturation) of the variable magnet plus or minus the magnetic flux (magnetic flux density) of the variable magnet. Is set as a value obtained by adding. If the positive maximum value of the magnetic flux amount of the variable magnet VMG is Φvarmax (assuming that the absolute value of the negative maximum value is equal to the positive maximum value) and the magnetic flux amount of the fixed magnet FMG is Φfix, the following equation is obtained.

  A magnetic current command Im * for obtaining the magnetic flux command Φ * is output from the magnetic current command 27 using the magnetic flux command Φ * as an input and the corresponding magnetizing current table 27 storing the magnetizing current.

Basically, since the magnetization direction of the magnet is the D axis, the magnetizing current command Im * is given to the D axis current command Id *. In the present embodiment, the D-axis current reference IdR that is the output from the current reference calculation unit 11 is corrected by the adder 14 with the D-axis magnetization current command difference ΔIdm * to obtain the D-axis current command Id *. Therefore, the subtractor 26 calculates the D-axis magnetization current command ΔIdm * by the following equation.

  It is also possible to adopt a configuration in which the magnetizing current Im * is directly given to the D-axis current command Id * when switching the magnetic flux.

  On the other hand, the magnetization request flag FCreq is at least momentarily switched (FCreq = 1) when a request for switching the magnetic flux is made. In order to reliably change the magnetic flux, the magnetization request flag FCreq is input to the minimum on-pulse device 28. The magnetization completion flag (= 1: during magnetization = 0: magnetization completion), which is an output, has a function of not being turned off (= 0) for a predetermined time once turned on (= 1). . When the input is on (= 1) for a predetermined time, the output is turned off at the same time as it is turned off.

  The switch 23 receives a magnetization completion flag, and outputs the output of the subtractor 26 when the magnetization is in progress (magnetization completion flag = 1), and outputs 0 when the magnetization is complete (magnetization completion flag = 0).

  Based on the DQ-axis current commands Id * and Iq * generated as described above, the voltage command calculation unit 10 generates DQ-axis voltage commands Vd * and Vq * including a current controller so that a current matching the command flows. To do.

  The DQ axis voltage commands Vd * and Vq * of the voltage command calculation unit 10 are converted into three-phase voltage commands Vu *, Vv * and Vw * by the coordinate conversion unit 5, and the PWM circuit 6 is converted by the three-phase voltage command. A gate signal is generated by PWM, and the inverter 4 is PWM-controlled.

  As described above, the coordinate conversion unit 7 converts the AC detection currents Iu and Iw of the current detector 2 into two-axis DQ axes to convert them into DQ-axis current detection values Id and Iq, and a voltage command calculation unit. Enter 10. Further, the pseudo-differentiator 8 obtains the inverter frequency ω1 from the signal from the speed detector 18. The voltage command calculation unit 10, the coordinate conversion units 5 and 7, and the PWM circuit 6 employ known techniques similar to those in the prior art.

  FIG. 13 shows an example of a timing chart of the operation of each signal. Here, although the protection signal is not raised (PROT = 0), the magnetization request flag is raised by the change of the operation state flag Run and the change of the magnetic flux command Φ *, and the magnetization completion flag for securing the predetermined time width is raised. The magnetization current command Im * has a value only during the period of the magnetization completion flag.

  With the above configuration, the present embodiment has the following operational effects. The variable magnetic flux motor 1 has a sudden characteristic change with respect to the magnetization caused by the inverter current as shown in the BH characteristic of FIG. For this reason, even if the same control is applied in practice, the same D axis and magnetic flux axis, which are likely to occur in position sensorless control, may not be exactly the same due to an axis deviation, a difference in current response, and individual motor differences. It is difficult to obtain the magnetic flux repeatedly. If the repetition accuracy of the magnetic flux is poor, the torque accuracy is deteriorated, which is not preferable.

  However, according to the variable magnetic flux drive system of the present embodiment, the variable magnet VMG is set to flow a magnetization current that is equal to or greater than the magnetization saturation region in the magnetization characteristics of the variable magnet VMG. The repeatability can be improved, thus ensuring torque accuracy and improving drive reliability.

  In addition, according to the variable magnetic flux drive system of the present embodiment, since the minimum time for flowing the magnetizing current is set, it does not end in the halfway magnetization state, thereby making the variable after the magnetization processing variable. Variations in the amount of magnetic flux can be suppressed and torque accuracy can be improved.

Next, returning to FIG. 6, the function of the temperature information processing unit 122 will be further described. The temperature information processing unit 122 can perform any one or more of the following controls. That is,
(a) When it is detected that the junction temperature of the semiconductor power element is higher than a certain temperature, the variable magnetic flux motor is not magnetized or the magnetizing current can be made smaller than usual. For this purpose, for example, the output of the magnetization request flag from the magnetization request generator 29 is prohibited. Alternatively, there is a method for prohibiting output of a command current from the correction unit 145.

  (b) When the junction temperature of the semiconductor power element reaches a certain temperature during the magnetization of the variable magnetic flux motor, the magnetization can be stopped or the magnetization current can be suppressed. Also in this case, there is a method of stopping the output of the command current from the correction unit 145.

  (c) As explained in FIG. 4, a value proportional to the semiconductor power element junction temperature is subtracted from the command current value, and the semiconductor power element junction temperature is subtracted from the element upper limit temperature (junction upper limit temperature). There may be provided means for adding a value obtained by integrating the values every predetermined time to the command current value.

  (d) Further, in addition to the above (a)-(c), a magnetizing current calculated from the inverter input DC voltage and the switching frequency may be passed. For example, in accordance with the output value of the subtracter 144 in FIG. 4, a control output is obtained in which a magnetizing current obtained in advance from the inverter input DC voltage and the switching frequency is supplied to the variable magnetic flux motor. For this purpose, appropriate control data previously obtained through experiments or the like is prepared in the memory in accordance with the inverter input DC voltage and the switching frequency. Then, control data corresponding to the output value of the subtracter 144 is read from the memory and used as current control data that lowers the temperature of the inverter. Thereby, the temperature rise of an inverter can be suppressed.

  (E) In addition to the above (a)-(c), the inverter input voltage and switching frequency may be lowered before changing the magnetic force of the variable magnetic flux motor. Before changing the magnetic force of the variable magnetic flux motor, it can be determined, for example, at the output timing of the magnetization request flag. For example, when the output value of the subtracter 144 in FIG. 4 is within a certain range, it can be determined in which region (safety region, warning region, dangerous region, etc.) the temperature of the semiconductor power element is. .

  Therefore, when in the warning area, before changing the magnetic force of the variable magnetic flux motor, in order to change the inverter input voltage, for example, the D-axis current reference IdR may be switched to a preset value. Thereby, the temperature fall of an inverter can be obtained. Further, the output voltage of the inverter or the frequency of the switching control signal in the PWM circuit 6 may be directly adjusted by the output from the temperature information processing unit 122.

  (F) Furthermore, in addition to the above (a) to (c), means for enhancing the cooling of the inverter before changing the magnetic force of the variable magnetic flux motor may be provided. Normally, the inverter 4 module is integrated with a heat radiating fin or a heat radiator. In some cases, the cooling liquid circulates through the radiator. The cooling liquid is cooled by a cooler such as a radiator, flows through the radiator, and absorbs heat. Therefore, for example, it can be realized by reducing the amount of liquid flowing through the radiator, further cooling the liquid, and pouring the cooled liquid into the radiator of the inverter.

  (G) Depending on the value of the output of the subtractor 144 in FIG. 4, a control output that causes other drive systems to drive the system before changing the magnetic force of the variable magnetic flux motor may be obtained. This control output is effectively used in, for example, a hybrid vehicle. For example, when the inverter temperature reaches a dangerous area while running with a variable magnetic flux motor as power, it is necessary to send a start signal to the engine fueled with gasoline. The control output from the temperature information processing unit 122 is used.

  (H) Further, after changing the magnetic force of the variable magnetic flux motor according to the value of the output of the subtractor 144 in FIG. 4, a control output that does not change the magnetic force of the variable magnetic flux motor for a certain period may be obtained. This takes into account the cooling period of the inverter. When control such as stopping the magnetization request flag or switching the command current Id to a preset value is performed by the control output from the temperature information processing unit 122, the control state is maintained for a certain period. Thereby, a sufficient cooling period of the inverter can be ensured.

  Note that the inverter 4 includes a plurality of semiconductor power elements that perform switching operation, and is provided with a temperature sensor for each element or a temperature sensor for an element that is in an on / off relationship. In this case, the average data of the detected temperature data from each temperature sensor may be processed as measured temperature data by the temperature information processing unit 122, or the highest detected temperature data may be processed as measured temperature data by the temperature information processing unit 122. May be processed.

  FIG. 14 shows an example of means for enhancing the cooling of the inverter 4. Normally, the inverter 4 module is integrated with a heat radiating fin or a heat radiator. Cooling liquid circulates through the radiator. The cooling liquid is cooled by a cooler such as the radiator 150 and flows through the radiator of the inverter 4 to absorb heat. Further, the cooling water flows around the motor 1 or the battery to cool it, returns to the radiator 150, and undergoes a cooling process. Thus, in order to enhance the cooling effect in the inverter 4, for example, it can be realized by reducing the amount of liquid flowing through the radiator 150, cooling the liquid more strongly, and pouring the cooled liquid into the radiator of the inverter 4.

  By providing the above function, it is possible to accurately estimate the temperature change at the junction of the semiconductor power element, and to obtain operational safety and reliability of the inverter and the variable magnetic flux motor.

  The above-described present invention is a technology effective in the field of equipment using a motor such as an electric vehicle, a hybrid vehicle, and a train.

  DESCRIPTION OF SYMBOLS 1 ... Variable magnetic flux motor, 2U, 2W ... Current detection device, 3 ... DC power supply, 4 ... Inverter, 5 ... Coordinate conversion part, 6 ... PMW circuit, 7 ... Coordinate conversion unit, 8 ... pseudo-differentiator, 10 ... voltage command calculation unit, 11 ... current reference calculation unit, 12 ... magnetic flux command calculation unit, 13 ... variable magnetic flux control unit, 14 ..Adder, 15 ... Gate command generation unit, 16 ... Operation command generation unit, 17 ... Protection determination unit, 29 ... Magnetization request generation unit, 111 ... Semiconductor power element, 112. ..Gate switch circuit, 121 ... temperature sensor, 122 ... temperature information processing unit, 124 ... adder, 130 ... modeling circuit, 141 ... proportional constant calculation unit, 142 ... Subtractor, 43... Integral constant calculation unit, 144... Subtractor, 145.

Claims (10)

Define a modeling circuit including the thermal resistance and heat capacity between the junction of the semiconductor power element and the on-chip sensor, input the direct output measured by the on-chip sensor to the modeling circuit, and output the modeling circuit Is an estimated loss, and this estimated loss is added to the direct output of the on-chip sensor to obtain estimated measured temperature data. A modeling circuit including a thermal resistance and a thermal capacity between the junction of the semiconductor power element and the on-chip sensor is defined, and a first value proportional to the junction temperature data of the semiconductor power element is obtained from the thermal resistance of the modeling circuit. Obtaining a second value obtained by integrating a value obtained by subtracting the junction temperature data of the semiconductor power element from preset upper limit temperature data using the heat capacity of the modeling circuit every predetermined time; A method for controlling an inverter, wherein a third value obtained by subtracting the second value from the first value is subtracted from an inverter command current value. A modeling circuit including a thermal resistance and a heat capacity between the junction of the semiconductor power element in the inverter and the on-chip sensor is defined, and the temperature information processing unit including the modeling circuit includes:
A proportionality constant calculator that obtains a first value proportional to the junction temperature data of the semiconductor power element from the thermal resistance of the modeling circuit;
An integration constant calculation unit that obtains a second value by integrating a value obtained by subtracting the junction temperature data of the semiconductor power element from preset upper limit temperature data every predetermined time using the heat capacity of the modeling circuit When,
A subtractor for obtaining a third value obtained by subtracting the second value from the first value;
A correction unit that subtracts the third value from the command current value for the inverter;
A control apparatus for a variable magnetic flux motor, characterized by comprising: When the temperature information processing unit monitors the third value and detects a state in which the junction temperature of the semiconductor power element is higher than a preset temperature, the variable magnetic flux motor is not magnetized. 4. The control device for a variable magnetic flux motor according to claim 3, wherein a control output for making the magnetizing current smaller than usual is obtained. The temperature information processing unit monitors the third value and stops the magnetization when the junction temperature of the semiconductor power element reaches a preset temperature during the magnetization of the variable magnetic flux motor. 4. The control device for a variable magnetic flux motor according to claim 3, wherein a control output for suppressing the magnetizing current is obtained. The temperature information processing unit obtains a control output for causing a magnetizing current obtained in advance from an inverter input DC voltage and a switching frequency to flow through the variable magnetic flux motor according to the third value. Control device for variable magnetic flux motor. 4. The variable magnetic flux motor according to claim 3, wherein the temperature information processing unit obtains a control output for lowering the inverter input voltage and the switching frequency before changing the magnetic force of the variable magnetic flux motor according to the third value. Control device. 4. The control of a variable magnetic flux motor according to claim 3, wherein the temperature information processing unit obtains a control output for enhancing cooling of the inverter before changing the magnetic force of the variable magnetic flux motor in accordance with the third value. apparatus. 4. The temperature information processing unit obtains a control output for causing another drive system to drive the system before changing the magnetic force of the variable magnetic flux motor according to the third value. Control device for variable magnetic flux motor. 4. The temperature information processing unit obtains a control output that does not change the magnetic force of the variable magnetic flux motor for a certain period of time after changing the magnetic force of the variable magnetic flux motor according to the third value. The control apparatus of the described variable magnetic flux motor.

Images