Dynamic optimization control method and system for thermal performance of traction converter

Abstract

The present invention relates to the field of power electronics technology, and discloses a method and system for dynamically optimizing the thermal performance of a traction converter to achieve dynamic optimization control of the thermal performance of a traction converter under normal operation and fault operation. The method includes: constructing a prediction model of an electrical performance control quantity and a prediction model of a thermal performance control quantity of a traction converter to be tested; constructing an electrical performance control objective function according to the prediction model of the electrical performance control quantity, and constructing a thermal performance control objective function based on dynamic weighting of a standard deviation coefficient according to the prediction model of the thermal performance control quantity; constructing a reward function for representing performance normalization based on the electrical performance control objective function and the thermal performance control objective function; determining a system level state combination set that meets preset conditions according to the operating state of the system; and adjusting the electrical and thermal performance of the traction converter to be tested according to the target system level state combination.

Description

Technical Field

[0001] The present invention relates to the field of rail transit power technology, and in particular to a method and system for dynamically optimizing the thermal performance of a traction converter. Background Art

[0002] As the core component system of high-speed trains, the traction drive system is mainly responsible for the task of driving the train to run at high speed and smoothly. The traction drive system is mainly composed of key equipment such as traction converters and traction motors. At present, the control strategies adopted by the traction drive systems of my country's high-speed trains are all based on normal operation to control the system, ignoring the differences in the actual execution effect of the equipment on the control target. Especially in the case of system failure, the differences caused by different devices, different types, and different positions in the system are not only obvious, but also more diverse; in addition, in the existing control strategies, only train speed and current tracking are used as control targets, lacking consideration of equipment reliability. The above shortcomings, on the one hand, will cause the high-speed train to be unable to cope with the remedial safe operation problem after the sudden failure of the system during operation, which will bring hidden dangers to the stable and safe operation of the train running under the failure; on the other hand, it will lead to low reliability level, short service life and high risk of failure of the key equipment of the system during actual operation, thereby increasing the maintenance cost.

[0003] Among them, the traction inverter, as the core equipment of the traction drive system, is responsible for the energy supply and conversion of the traction drive train. According to relevant statistical results, open circuit failure of power devices is the main source of traction inverter failure, which is one of the high-frequency maintenance devices of the train. Once the traction inverter fails, it will affect the normal operation of the system, and in severe cases, the train will be forced to stop. It can be seen that the reliability of the traction inverter directly affects the safe operation level of the traction drive system and even the entire vehicle.

[0004] According to the experimental results, in addition to over-electric stress and other extreme abnormal conditions, thermal cycle stress accumulation and impact are the main causes of failure of power module devices. It should be noted that with the popularization and application of traction converter integration and modularization, several power devices are integrated into one power module. Therefore, if a power module is damaged due to thermal stress, the entire power module or even the traction converter equipment will be replaced and repaired. However, the power module also includes a large number of other normal power devices with a long remaining life. In fact, since the thermal stress of the power module is difficult to simply add to the traditional closed-loop control strategy, the thermal stress condition of the traction converter is ignored, especially in the application of multi-level traction converters, the thermal stress imbalance problem between the power devices in the power module of the traction converter is more prominent. At present, the reliability improvement technology of traction converters based on active temperature management has attracted widespread attention, but related research is still in its infancy. It mainly focuses on improving the electrothermal performance of the converter around a single power device or using the symmetry of the converter topology under normal system operation, and does not involve complex situations such as the loss of symmetry of the converter topology due to system failure and the phase current of the converter is no longer symmetrical and balanced.

[0005] Therefore, how to achieve dynamic optimization control of the thermal performance of the traction converter under normal operation and fault operation conditions becomes an urgent problem to be solved. Summary of the invention

[0006] The purpose of the present invention is to disclose a method and system for dynamically optimizing the thermal performance of a traction inverter, so as to realize dynamic optimization control of the thermal performance of the traction inverter under normal operation and fault operation conditions by intelligently regulating the working state of each power module in the traction inverter.

[0007] To achieve the above object, the present invention provides a method for dynamically optimizing the thermal performance of a traction converter, comprising:

[0008] S1: Constructing a prediction model for the electrical performance control quantity and a prediction model for the thermal performance control quantity of the traction converter to be tested;

[0009] S2: constructing an electrical performance control objective function according to the prediction model of the electrical performance control quantity, and constructing a thermal performance control objective function based on a dynamic weighting of a standard deviation coefficient according to the prediction model of the thermal performance control quantity;

[0010] S3: constructing a reward function for representing performance normalization based on the electrical performance control objective function and the thermal performance control objective function;

[0011] S4: Determine a system level state combination set that meets preset conditions according to the operating state of the system;

[0012] S5: adjusting the electrothermal performance of the traction converter to be tested according to the target system level state combination, wherein the target system level state combination is a system level state combination that maximizes the function value output by the reward function in the system level state combination set.

[0013] As a general inventive concept, the present invention also provides a traction inverter thermal performance dynamic optimization control system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the corresponding steps of the above method when executing the computer program.

[0014] The present invention has the following beneficial effects:

[0015] The method for dynamically optimizing the thermal performance of a traction converter provided by the present invention can intelligently control the heat generated by the power module in the traction converter and the current path and voltage in the traction converter by adjusting the electrothermal performance of the traction converter to be tested, thereby optimizing the electrothermal performance of the traction converter to be tested. The method is easy to implement, does not require additional hardware equipment, can extend the overall electrothermal performance and service life of the traction converter, can improve the reliability of train operation, and reduce equipment maintenance costs.

[0016] The present invention will be further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings constituting a part of this application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a topological diagram of a three-level traction inverter according to an embodiment of the present invention;

[0019] Figure 2 It is a block diagram of the overall control principle of a three-level traction inverter system for a permanent magnet train according to an embodiment of the present invention;

[0020] Figure 3 is a flow chart of a method for dynamically optimizing thermal performance of a traction converter according to an embodiment of the present invention;

[0021] Figure 4 It is the dynamic weight coefficient in the control objective function of the thermal control quantity of the embodiment of the present invention when the system is in normal operation and adopts the thermal performance dynamic optimization control strategy, the thermal performance optimization control strategy and the traditional finite set model predictive control strategy. Schematic diagram of

[0022] Figure 5It is a schematic diagram of the mean value of the overall junction temperature of the three power modules and the mean value and variance of the junction temperature variance of each power module under normal system operation in an embodiment of the present invention, using a thermal performance dynamic optimization control strategy, a thermal performance optimization control strategy and a traditional finite set model predictive control strategy;

[0023] Figure 6 It is a schematic diagram of the system stator three-phase current and train running speed under normal system operation in an embodiment of the present invention, using a thermal performance dynamic optimization control strategy, a thermal performance optimization control strategy and a traditional finite set model predictive control strategy;

[0024] Figure 7 It is the first power device IT of the U phase of the traction inverter in the embodiment of the present invention. 11 When a fault occurs, the dynamic weight coefficient in the control objective function of the thermal control quantity using the thermal performance dynamic optimization control strategy, the thermal performance optimization control strategy and the traditional finite set model predictive control strategy is Schematic diagram of

[0025] Figure 8 It is the first power device IT of the U phase of the traction inverter in the embodiment of the present invention. 11 When a fault occurs, the mean junction temperature of the three power modules as a whole and the mean and variance of the junction temperature variance of each power module using the thermal performance dynamic optimization control strategy, the thermal performance optimization control strategy and the traditional finite set model predictive control strategy are shown;

[0026] Fig. 9 It is the first power device IT of the U phase of the traction inverter in the embodiment of the present invention. 11 Schematic diagram of system stator three-phase current and train running speed using thermal performance dynamic optimization control strategy, thermal performance optimization control strategy and traditional finite set model predictive control strategy when a fault occurs;

[0027] Fig.10 It is the U-phase No. 2 power device IT of the traction inverter in the embodiment of the present invention. 12 When a fault occurs, the dynamic weight coefficient in the control objective function of the thermal control quantity using the thermal performance dynamic optimization control strategy, the thermal performance optimization control strategy and the traditional finite set model predictive control strategy is Schematic diagram of

[0028] Fig.11 It is the U-phase No. 2 power device IT of the traction inverter in the embodiment of the present invention. 12 When a fault occurs, the mean junction temperature of the three power modules as a whole and the mean and variance of the junction temperature variance of each power module using the thermal performance dynamic optimization control strategy, the thermal performance optimization control strategy and the traditional finite set model predictive control strategy are shown;

[0029] Fig.12It is the U-phase No. 2 power device IT of the traction inverter in the embodiment of the present invention. 12 Schematic diagram of system stator three-phase current and train running speed using thermal performance dynamic optimization control strategy, thermal performance optimization control strategy and traditional finite set model predictive control strategy when a fault occurs;

[0030] Fig.13 The U-phase clamping diode D of the traction inverter in the embodiment of the present invention is 11 When a fault occurs, the dynamic weight coefficient in the control objective function of the thermal control quantity using the thermal performance dynamic optimization control strategy, the thermal performance optimization control strategy and the traditional finite set model predictive control strategy is Schematic diagram of

[0031] Fig.14 The U-phase clamping diode D of the traction inverter in the embodiment of the present invention is 11 When a fault occurs, the mean junction temperature of the three power modules as a whole and the mean and variance of the junction temperature variance of each power module using the thermal performance dynamic optimization control strategy, the thermal performance optimization control strategy and the traditional finite set model predictive control strategy are shown;

[0032] Fig.15 The U-phase clamping diode D of the traction inverter in the embodiment of the present invention is 11 Schematic diagram of system stator three-phase current and train running speed using thermal performance dynamic optimization control strategy, thermal performance optimization control strategy and traditional finite set model predictive control strategy when a fault occurs. DETAILED DESCRIPTION

[0033] The embodiments of the present invention are described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways as defined and covered by the claims.

[0034] It should be understood that the method and system for dynamically optimizing the thermal performance of a traction converter in the embodiments of the present application can be applied to rail transit application scenarios, such as high-speed trains, EMUs, or subways, etc. This is only an example and not a limitation.

[0035] Example 1

[0036] This embodiment refers to a three-level traction converter system of a certain type of permanent magnet high-speed train. The traction converter can be further divided into a single-phase three-level rectifier and a three-phase three-level inverter. The specific topology is as follows: Figure 1As shown. This embodiment will take the dynamic optimization control of thermal performance of a three-phase three-level traction inverter as an example for explanation. The traction three-level inverter system adopts a speed-current dual closed-loop control structure, and the outer control loop is a speed loop. According to the speed feedback, the given value of the system electromagnetic torque can be obtained through a lookup table; then, a full-speed embedded permanent magnet synchronous motor control strategy of maximum torque current ratio (MTPA) and weak magnetic control based on motor stator d-axis current compensation is adopted to obtain the given values ​​of the motor stator d-axis current and stator q-axis current, and this is used as the given input of the current inner loop control strategy; finally, the thermal performance dynamic optimization control strategy selects the system level state combination that maximizes the reward function value as the system control command output, so as to realize the tracking of the given values ​​of the motor stator d and q-axis currents and the intelligent regulation of the power module heat. The control principle block diagram is shown as follows: Figure 2 shown.

[0037] In this embodiment, main parameters of a three-level traction inverter system of a certain type of permanent magnet high-speed train are shown in Table 1.

[0038] Table 1 Main parameters of permanent magnet train three-level traction inverter system

[0039]

[0040]

[0041] like Figure 3 As shown, a method for dynamically optimizing the thermal performance of a traction converter provided in an embodiment of the present application includes the following steps.

[0042] S1: Construct a prediction model for the electrical performance control quantity and a prediction model for the thermal performance control quantity of the traction converter to be tested.

[0043] It should be noted that if Figure 2 As shown, the dynamic optimization control method for thermal performance of traction converter in this embodiment is in the current inner loop control link in the speed-current dual closed-loop control structure of traction converter. Therefore, the system electrical performance control quantity in this embodiment is the stator d-axis current and stator q-axis current of the embedded permanent magnet synchronous motor; correspondingly, the system thermal performance control quantity in this embodiment is the energy consumption of each power device that can be equivalent to the device temperature / junction temperature.

[0044] S2: constructing an electrical performance control objective function according to the prediction model of the electrical performance control quantity, and constructing a thermal performance control objective function based on dynamic weighting of standard deviation coefficients according to the prediction model of the thermal performance control quantity.

[0045] S3: Based on the electrical performance control objective function and the thermal performance control objective function, construct a reward function for representing performance normalization.

[0046] S4: According to the operating state of the system, determine a system level state combination set that meets the preset conditions.

[0047] In this step, the system level state combination set that meets the preset condition may represent all available level state combinations in the system.

[0048] S5: adjusting the electrothermal performance of the traction converter to be tested according to the target system level state combination, wherein the target system level state combination is a system level state combination that maximizes the function value output by the reward function in the system level state combination set.

[0049] The above-mentioned dynamic optimization control method for the thermal performance of the traction inverter can intelligently control the heat generated by the power module in the traction inverter and the current path and voltage in the traction inverter by adjusting the electrothermal performance of the traction inverter to be tested, thereby realizing the optimization control of the electrothermal performance of the traction inverter to be tested. The method is easy to implement, does not require additional hardware equipment, can extend the overall electrothermal performance and service life of the traction inverter, can improve the reliability level of train operation, and reduce equipment maintenance costs.

[0050] Optionally, adjusting the electrothermal performance of the traction converter to be measured according to the target system level state combination includes:

[0051] When the traction converter to be tested is in a normal operating state or a fault state, the overall temperature of the traction converter to be tested is reduced, and the temperature differences of the bridge arms tend to be consistent, or;

[0052] Reduce the average temperature of a single power module to make the temperature differences between the power devices in the module more consistent;

[0053] The fault state is used to indicate that an open circuit fault occurs in a power device or a clamping diode in a power module of the traction converter.

[0054] In this optional embodiment, making the temperature difference between the power devices in the module tend to be consistent means reducing the average temperature difference of the power devices in the module. By lowering the overall temperature of the traction converter to be tested, and the temperature difference of each bridge arm tends to be consistent, or lowering the average temperature of a single power module, the average temperature difference of each power device in the module is reduced to a preset threshold. For example, the heat generated by the power module in the traction converter and the current path and voltage in the traction converter can be regulated to adjust the temperature in the module, so that the system thermal performance can be enhanced when the traction converter is in normal operation; when the traction converter is in a faulty operation, the system thermal performance can be improved.

[0055] Optionally, the electrical performance control quantity of the traction inverter includes but is not limited to phase current, DC voltage, average power, or transient speed; the thermal performance control quantity of the traction inverter includes but is not limited to device junction temperature or device energy consumption.

[0056] It should be noted that, in this optional implementation, the electrical performance of the traction converter refers to the properties and functions of the electrical parameters (including current, voltage, power, speed, etc.) of the traction converter system, which can be quantitatively evaluated by calculation formulas such as total harmonic distortion, ripple factor, steady-state error, tracking error, etc. The system electrical performance control quantity refers to the electrical variables (including phase current, DC voltage, average power, transient speed, etc.) that can be measured, estimated, and regulated in the traction converter system; the thermal performance of the traction converter refers to the properties and functions of the thermal field parameters (temperature, energy consumption, heat dissipation, etc.) of each power device, power module, and converter in the traction converter system, which can be quantitatively evaluated by calculation formulas such as average value, maximum value, transient value, variance value, etc. The system thermal performance control quantity refers to the thermal field variables (including device junction temperature, device energy consumption, etc.) that can be measured, estimated, and regulated in the traction converter system. This is only an example and is not limited.

[0057] Optionally, the S1 includes:

[0058] S11: Construct a prediction model for the electrical performance control quantity of the traction converter to be tested, expressed as:

[0059] x[n+1]=A[n]x[n]+B(S[n])u[n]+C[n]; (1)

[0060] Where x[n+1] is the state vector of the system electrical performance control quantity predicted in the [n+1]th sampling period, [n] is the [n]th sampling period, x[n] is the state vector of the system electrical performance control quantity in the [n]th sampling period, and x∈{x1,x2,…,x p ,…,x P}, x p is the pth state variable in the state space equation related to the control quantity of the system electrical performance, x p The value of is the sampling value of the system electrical performance control quantity in the [n]th sampling period, P is the total number of state variables in the state space equation related to the electrical performance control quantity, A is the time-varying parameter matrix of the state vector in the state space equation related to the system electrical performance control quantity, B(S[n]) is the time-varying parameter matrix of the input vector in the state space equation related to the system electrical performance control quantity, and its expression is related to S[n]. S[n] is the set of all combinations of the level states of each phase bridge arm of the traction converter in the [n]th sampling period, S[n]=[S 1 [n],S 2 [n],…,Sd [n],…,S D [n]],S d [n] represents the level state of the d-phase bridge arm of the traction converter in the [n]th sampling period, where the level state of the bridge arm is composed of a number of discrete values ​​and is determined by the topological structure of the traction converter, and D is the total number of bridge arms of the traction converter; u[n] is the input vector of the system in the [n]th sampling period, where the representation of u[n] satisfies u[n]=[u1[n],u2[n],…,u q [n],…,u Q [n]] T ,u q [n] is the qth input variable, T represents the transpose of the matrix, Q is the total number of input variables; C is the time-varying parameter matrix composed of non-vector terms in the state space equation related to the control quantity of the system electrical performance.

[0061] In this embodiment, the traction three-level inverter has three-phase bridge arms, and each phase bridge arm is composed of four power modules, such as Figure 1 As shown, they are the U-phase bridge arm, the V-phase bridge arm and the W-phase bridge arm. Under normal circumstances, each bridge arm has three level states. Therefore, in this embodiment, D=3, S 1 , S 2 and S 3 The level states of the U-phase bridge arm, the V-phase bridge arm and the W-phase bridge arm are respectively corresponded. Specifically, in this embodiment, the level state S of the d-th phase bridge arm is d , which can be expressed as:

[0062]

[0063] In the formula, and They respectively represent the control signals that determine the on / off state of the four power devices in each bridge arm. "1" indicates that the control power device is in the on state, and "0" indicates that the control power device is in the off state. The corresponding relationship is as follows: Figure 1 shown.

[0064] like Figure 2 As shown, in this embodiment, the system electrical performance control quantity is the motor stator d, q axis current i d and i q . Specifically, x=[i d ,i q ] T , the total number of state variables in the state space equation related to the electrical performance control quantity P = 2, the total number of input variables Q = 1, u = [u cd ], the time-varying parameter matrices in formula (1) are:

[0065]

[0066]

[0067]

[0068] Where, L d and L q are the inductances of the motor stator d-axis and q-axis, t s is the system sampling period, ω m [n] is the motor mechanical angular velocity, θ e [n] is the electrical angle of the system in the [n]th system sampling period, which can be the mechanical angle θ measured by the system speed sensor m [n] multiplied by the number of pole pairs n of the permanent magnet synchronous motor p Calculated, such as Figure 2 shown.

[0069] Step 12: Construct a prediction model for the thermal performance control quantity of the traction converter to be tested, expressed as:

[0070]

[0071] In the formula, is the predicted total energy consumption of the jth power device in the kth power module of the traction converter in the [n+1]th sampling period, k = 1, 2, ..., K, K is the total number of power modules in the traction converter, j = 1, 2, ..., J k , J k is the total number of power devices in the kth power module; S k [n] represents the level state in the kth power module of the traction converter in the [nth]th sampling period; i k [n+1] is the predicted value of the current flowing through the kth power module of the traction converter in the [n+1]th sampling period, which can be calculated based on S11. kj [n] is the junction temperature of the jth power device in the kth power module in the [n]th sampling period; and are the conduction loss and switching loss of the jth power device in the kth power module of the traction converter, respectively. Their values ​​are calculated based on the junction temperature T kj [n] and the current prediction value i k [n+1] determined; and They are functions for determining whether the j-th power device in the k-th power module generates conduction loss and switching loss, respectively.

[0072] In this embodiment, i k[n+1] can be calculated by formula (1) as follows: the stator d-axis current i of the permanent magnet synchronous motor in the dq coordinate system during the [n+1]th system sampling period d [n+1] and stator q-axis current i q [n+1] is calculated by Park inverse transform and Clark inverse transform. Specifically, the formula can be expressed as:

[0073]

[0074] In this embodiment, the expression on the right side of formula (3) is and The expression can be a fitting function, and the fitting data comes from the user data manual of the power module manufacturer.

[0075] In this embodiment, the function for determining whether conduction loss and switching loss occur is and The expression of can be consistent with the description in the prior art and will not be repeated here.

[0076] In this embodiment, Figure 2 The four power devices of one phase bridge arm in the embodiment are regarded as a power module. Specifically, K=3, J k are both 4; therefore, in this embodiment, S k =S d Specifically, in this embodiment, the level state S of the kth power module k , which can be expressed as:

[0077]

[0078] In the formula, s k1 ,s k2 ,s k3 and k4 1 and 0 respectively represent control signals for determining the on / off states of the four power devices in each power module in this embodiment. "1" indicates that the control power device is in the on state, and "0" indicates that the control power device is in the off state. The corresponding relationship is as follows: Figure 1 shown.

[0079] Optionally, the S2 includes:

[0080] S21: constructing an electrical performance control objective function according to the prediction model of the electrical performance control quantity, expressed as:

[0081]

[0082] In the formula, g electric(l)[n] is the value of the control objective function of the system electrical performance under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [nth]th system sampling period, where l = 1, 2, ..., L, L is the total number of all available combinations of the level states of each phase bridge arm of the traction converter; x p_ref [n] is the reference value of the pth state variable in the system electrical performance control quantity in the [nth]th system sampling period; x p [n+1](S(l)[n]) is the predicted value of the pth state variable in the system electrical performance control quantity in the [n+1]th system sampling period under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [n]th system sampling period obtained according to S11. is the weight coefficient of the pth state variable in the system electrical performance control quantity in the [n]th system sampling period.

[0083] It should be noted that, in this embodiment, the system electrical performance control objective function is:

[0084]

[0085] In the formula, i d_ref [n-1] and i q_ref [n-1] are the given values ​​of the stator d-axis current and stator q-axis current of the permanent magnet synchronous motor in the dq coordinate system in the [n-1]th system sampling period, which are calculated by the system control outer loop. Figure 2 shown.

[0086] It should be noted that, in this embodiment, λ1 electric =0.3,λ1 electric =0.7.

[0087] S22: Constructing a thermal performance control objective function based on the dynamic weighting of the standard deviation coefficient according to the prediction model of the thermal performance control amount, expressed as:

[0088]

[0089]

[0090]

[0091] In the formula, g thermal (l)[n] is the value of the system thermal performance control objective function under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [nth]th system sampling period. k () is the thermal performance control objective function of the kth power module of the traction converter based on the energy consumption variance. is the switching cycle t of the jth power device in the kth power module of the traction converter in the [n]th system sampling cycle before the [n]th system sampling cycle sw Total energy consumption in s is the system sampling period, J k ′ is the total number of power devices that can work normally in the kth power module of the traction converter. is the weight coefficient in the control objective function of the thermal performance control quantity of the kth power module of the traction converter in the [n]th system sampling period, N thermal The sampling period window size of the historical data used to calculate the weight coefficient of the system thermal performance control quantity is determined by the heat dissipation time constant τ of the power module. thermal With the system sampling period t s The ratio of is rounded to get, τ thermal The size of mod(n,N thermal ) is the dividend n, N thermal Modulo operation as divisor. T avg_k [n] is the average value of the sum of the junction temperatures of the normal power devices in the kth power module in the [nth]th system sampling period. kj [n] is the flag bit of the operating status of the jth power device in the kth power module in the [nth]th system sampling cycle. When the power device can work normally, flag kj [n] = 1; when a power device fails and cannot work normally due to the fault (current cannot flow through), the corresponding power device operating status flag is kj [n] = 0. The weight coefficient in the control objective function of the thermal performance control quantity in the embodiment of the present application is a dynamic thermal weight coefficient. In practice, after a fault occurs, the thermal performance of different power modules changes differently. Therefore, the dynamic thermal weight coefficient in the embodiment of the present application can achieve a better thermal balance effect.

[0092] It should be noted that, in this embodiment, the heat dissipation time constant τ of the power module thermal The time required for the transient thermal impedance of the thermal impedance curve of the power module to rise to 90% is adopted. Specifically, in this embodiment,

[0093] In this embodiment, t s =40μs. This is just an example and is not a limitation.

[0094] In this embodiment, the junction temperature T kj[n] Online estimation is performed using a thermal impedance equivalent network, and its expression is consistent with the description in the prior art and will not be repeated here.

[0095] It should be noted that, in this embodiment, an open circuit failure of a single power device and a single clamping diode of a traction inverter is taken as an example. Specifically, when the outer power device of the kth power module (i.e. Figure 2 Medium Power Devices IT k1 and IT k4 ) When a fault occurs, J k ′=3;When the inner power device of the kth power module (i.e. Figure 2 Medium Power Devices IT k2 and IT k3 ) When a fault occurs, J k ′=2;When the clamping diode of the kth power module (i.e. Figure 2 Middle D k1 and D k2 ) When a fault occurs, J k ′=4.

[0096] It should be noted that, in this embodiment, when a certain outer power device (i.e. Figure 2 Medium Power Devices IT k1 and IT k4 )When a fault occurs, flag k1 =0 or flag k4 =0; when a certain inner power device of the kth power module (i.e. Figure 2 Medium Power Devices IT k2 and IT k3 ) fails, although the adjacent outer power device does not fail, no current will flow through it, that is, it is in a pseudo-fault state. Accordingly, flag k1 =0, flag k2 =0 or flag k3 =0, flag k4 =0; when the clamping diode of the kth power module (i.e. Figure 2 Middle D k1 and D k2 ) When a fault occurs, all power devices of all power modules are working normally, that is, flag kj =1.

[0097] In this embodiment, the traction inverter U phase No. 1 power device IT k1 , traction inverter U phase No. 2 power device IT k2 and traction inverter U phase clamping diode D 11 Specifically, the power device IT k1 When a fault occurs, J′1=3, f11 =0, the flag bits of the other power devices’ operating status are all 1; the power device IT k2 When a fault occurs, J′1=2, flag 11 =0 and flag 12 =0, the flag bits of the operating status of the other power devices are all 1; the clamping diode D 11 When a fault occurs, J′1=4, and the flag bits of the operating status of all power devices are 1.

[0098] Optionally, the S3 includes:

[0099] Construct a reward function to represent the performance normalization, expressed as:

[0100]

[0101] Where R(S(l)) is the value of the reward function based on performance normalization under the action of the level state combination S(l) of each phase arm of the lth group of traction converters, α electri is the normalized control objective function G of the system electrical performance control quantity electri The weight coefficient, G electric (S(l)) is the value of the normalized control objective function of the system electrical performance control quantity under the action of the level state combination S(l) of each phase bridge arm of the lth group of traction converters. thermal is the normalized control objective function G of the system thermal performance control quantity thermal The weight coefficient, G thermal (S(l)) is the value of the normalized control objective function of the system thermal performance control quantity under the action of the combination S(l) of the level states of the bridge arms of each phase of the traction converter of the lth group. min{} represents the value of the control objective function corresponding to the minimum of the system electrical performance or thermal performance control objective function selected from all available combinations of the level states of the bridge arms of each phase of the traction converter.

[0102] In this embodiment, α electric =1,α thermal =1.2. This is just an example and not a limitation.

[0103] In this embodiment, the normalized electrical performance control objective function is:

[0104]

[0105] It should be noted that, in this embodiment, the normalized thermal performance control objective function satisfies the following relationship:

[0106]

[0107] In the formula, The relationship between S(l) and can be obtained from equation (3).

[0108] Optionally, in S4, the system level state combination set is expressed as:

[0109]

[0110] In the formula, S(l normal ) is the number of all available traction converter bridge arm level state combinations under normal system operation, and its number is the same as the number of all possible traction converter bridge arm level state combinations under normal system operation. normal =1,2,...,L normal , L normal It is the total number of all available combinations of bridge arm level states of each phase of the traction converter under normal system operation. is the combination of the bridge arm level states of all available traction converters when the system has the mth type of fault. is the total number of all available combinations of bridge arm level states of each phase of the traction converter when the mth type of fault occurs in the system; m = 1, 2, ..., M, M is the total number of all possible open circuit fault types of power devices in the system.

[0111] As a feasible example, in this embodiment, the traction inverter is a three-phase neutral point clamped three-level topology structure, so L normal =3 3 =27;

[0112] In this embodiment, the traction inverter U phase No. 1 power device IT 11 , traction inverter U phase No. 2 power device IT 12 and traction inverter U phase clamping diode D 11 Take the open circuit fault as an example, therefore, M=3.

[0113] In this embodiment, the traction inverter U phase No. 1 power device IT 11 A fault occurs and the current direction is such that it can flow through the power device IT 11 When L 1 fault =2*3 2 =18; Traction inverter U phase No. 2 power device IT 12 A fault occurs and the current direction is such that it can flow through the power device IT 12 hour, Traction inverter U phase clamping diode D 11 A fault occurs and the current direction is such that it can flow through the clamping diode IT 11 hour,

[0114] Optionally, the S5 includes:

[0115] S51: Taking the maximum reward function value as the optimization goal, a one-step optimized reward function is established, which is expressed as:

[0116]

[0117] Where R(S(H)) represents the value of the reward function corresponding to the maximum reward function based on performance normalization under the action of the Hth group of traction converter bridge arm level state combination S(H), H∈{1,2,...,l,...,L}; max{} represents the value of the reward function corresponding to the maximum reward function selected from all available combinations of the bridge arm level states of the traction converter phases.

[0118] S52: The combination S(H) of the bridge arm level states of each phase of the Hth group of traction converters is used as the system control command output within [n] sampling cycles to control the on / off state of each power device, thereby determining the heat generated by the power devices in the power module and the current path and voltage in the traction converter, so as to achieve dynamic optimization control of the thermal performance of the system when it is in normal operation or fault state.

[0119] In the nth system sampling cycle, i1[n], i2[n], i3[n], θ are obtained through system sensor sampling. e [n] and ω m [n]; further, the system outer loop control strategy can be used to obtain i d_ref [n],i q_ref [n] and i d_ref [n-1], i d_ref [n-1].

[0120] Specifically, in this embodiment, when the system is running at a certain stable speed, the thermal performance dynamic optimization control strategy and the thermal performance optimization control strategy (corresponding to the thermal weight coefficient ) and the dynamic weight coefficients in the control objective function of the thermal control quantity of the traditional finite set model predictive control strategy like Figure 4 As shown; the mean value of the junction temperature of the three power modules as a whole and the mean and variance of the junction temperature variance of each power module, as shown Figure 5 As shown; the system stator three-phase current and train running speed, such as Figure 6 shown.

[0121] Depend on Figure 4-Figure 6It can be seen that under normal system operation, compared with the thermal performance optimization control strategy with a fixed thermal weight coefficient, the thermal performance dynamic optimization control strategy with a dynamic thermal weight coefficient proposed in this patent has a more uniform thermal field distribution of each power module of the traction converter system, and the junction temperature and its fluctuation between each power device in the power module are more consistent. Compared with the traction converter system of the traditional finite set model predictive control strategy, the control method of this patent can greatly improve / enhance the thermal performance of the system while sacrificing less system electrical performance. According to relevant research results, when the thermal field distribution of each power module is more uniform, and the junction temperature and its fluctuation between each power device in the power module are more consistent, the life consumption of each power module will also be similar, and the overall service life of the traction converter will effectively avoid the "short board of the barrel" effect, thereby achieving the extension of its overall service life.

[0122] Specifically, in this embodiment, when the system is running at a certain stable speed, the first power device IT of the traction inverter U phase k1 Before and after a fault occurs, the thermal performance dynamic optimization control strategy and thermal performance optimization control strategy (corresponding to the thermal weight coefficient ) and the dynamic weight coefficients in the control objective function of the thermal control quantity of the traditional finite set model predictive control strategy like Figure 7 As shown; the mean value of the junction temperature of the three power modules as a whole and the mean and variance of the junction temperature variance of each power module, as shown Figure 8 As shown; the system stator three-phase current and train running speed, such as Fig. 9 shown.

[0123] Specifically, in this embodiment, when the system is running at a certain stable speed, the traction inverter U phase No. 2 power device IT k2 Before and after a fault occurs, the thermal performance dynamic optimization control strategy and thermal performance optimization control strategy (corresponding to the thermal weight coefficient ) and the dynamic weight coefficients in the control objective function of the thermal control quantity of the traditional finite set model predictive control strategy like Fig.10 As shown; the mean value of the junction temperature of the three power modules as a whole and the mean and variance of the junction temperature variance of each power module, as shown Fig.11 As shown; the system stator three-phase current and train running speed, such as Fig.12 shown.

[0124] Specifically, in this embodiment, when the system is running at a certain stable speed, the clamping diode D of the U phase of the traction inverter is 11 Before and after a fault occurs, the thermal performance dynamic optimization control strategy and thermal performance optimization control strategy (corresponding to the thermal weight coefficient ) and the dynamic weight coefficients in the control objective function of the thermal control quantity of the traditional finite set model predictive control strategy like Fig.13 As shown; the mean value of the junction temperature of the three power modules as a whole and the mean and variance of the junction temperature variance of each power module, as shown Fig.14 As shown; the system stator three-phase current and train running speed, such as Fig.15 shown.

[0125] By contrast Figure 4-Figure 6 and Figure 7-Figure 15 It can be seen that after the system fails, the thermal performance of the system deteriorates to varying degrees. The thermal field distribution of each bridge arm / power module of the traction converter deteriorates significantly. Not only the power devices in the faulty bridge arm / power module that can still work normally, but also the junction temperature and fluctuation differences of the power devices in other power modules increase. Specifically, when the system fails, Figure 7-Figure 15 It can be seen that compared with the thermal performance optimization control strategy with a fixed thermal weight coefficient, the thermal performance dynamic optimization control strategy with a dynamic thermal weight coefficient proposed in this patent can not only reduce the junction temperature and its fluctuation between the power devices in each power module after the fault, reducing the difference by more than 46%, but also make the thermal field distribution of each power module in the traction converter system more uniform. In addition, compared with the traction converter system of the traditional finite set model predictive control strategy, the control method of this patent can greatly improve the thermal performance of the system while sacrificing less system electrical performance, and can reduce the overall junction temperature average of the traction converter by about 0.5°C. Therefore, the thermal performance dynamic optimization control strategy with a dynamic thermal weight coefficient proposed in this application can more effectively improve the thermal performance under system faults (including more uniform thermal field distribution of each power module, more consistent junction temperature and its fluctuation between the power devices in the power module), and avoid the occurrence of secondary faults to a greater extent, thereby achieving the improvement of the service capability and reliability of the traction converter under fault operation.

[0126] It is worth noting that the above formulas are only one type of relational expressions, which are only used as examples and do not limit the calculation method.

[0127] In summary, the dynamic optimization control method for thermal performance of traction converter provided by this application completes the regulation of the heat generated by the power module and the current path and voltage in the traction converter, and realizes the dynamic optimization control of thermal performance of traction converter that integrates the enhancement of thermal performance of traction converter under normal system operation and the improvement of thermal performance of traction converter under system fault operation, thereby effectively extending the overall service life of traction converter, improving the service capacity and reliability of traction converter under fault, improving the reliability level of train operation, and reducing equipment maintenance costs. In addition, it is expected to become a leading key reliability technology with important theoretical research value and engineering application potential.

[0128] Example 2

[0129] Corresponding to the above method embodiment, this embodiment provides a traction inverter thermal performance dynamic optimization control system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the steps of the above method are implemented when the processor executes the computer program.

[0130] In summary, the method and system respectively disclosed in the above two embodiments of the present invention realize the dynamic optimization control of the thermal performance of the traction converter, which integrates the enhancement of the thermal performance of the traction converter under normal system operation and the improvement of the thermal performance of the traction converter under system fault operation, by intelligently regulating the heat generated by the power module in the traction converter and the current path and voltage in the traction converter. The method is easy to implement and does not require additional hardware equipment. It is of great significance to extend the service life of the traction converter, improve the service capacity, improve the reliability level of train operation, and reduce equipment maintenance costs.

[0131] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for dynamically optimizing the thermal performance of a traction converter, characterized in that: include: S1: Constructing a prediction model for the electrical performance control quantity and a prediction model for the thermal performance control quantity of the traction converter to be tested; S2: constructing an electrical performance control objective function according to the prediction model of the electrical performance control quantity, and constructing a thermal performance control objective function based on a dynamic weighting of a standard deviation coefficient according to the prediction model of the thermal performance control quantity; S3: constructing a reward function for representing performance normalization based on the electrical performance control objective function and the thermal performance control objective function; S4: Determine a system level state combination set that meets preset conditions according to the operating state of the system; S5: adjusting the electric and thermal performance of the traction converter to be tested according to the target system level state combination, wherein the target system level state combination is a system level state combination that maximizes the function value output by the reward function in the system level state combination set; The S1 includes: S11: Construct a prediction model for the electrical performance control quantity of the traction converter to be tested, expressed as: x[n+1]=A[n]x[n]+B(S[n])u[n]+C[n]; Where x[n+1] is the state vector of the system electrical performance control quantity predicted in the [n+1]th sampling period, [n] is the [n]th sampling period, x[n] is the state vector of the system electrical performance control quantity in the [n]th sampling period, and x∈{x1,x2,…,x p ,…,x P }, x p is the pth state variable in the state space equation related to the control quantity of the system electrical performance, x p The value of is the sampling value of the system electrical performance control quantity in the [n]th sampling period, P is the total number of state variables in the state space equation related to the electrical performance control quantity, A is the time-varying parameter matrix of the state vector in the state space equation related to the system electrical performance control quantity, B(S[n]) is the time-varying parameter matrix of the input vector in the state space equation related to the system electrical performance control quantity, and its expression is related to S[n]. S[n] is the set of all combinations of the level states of each phase bridge arm of the traction converter in the [n]th sampling period, S[n]=[S 1 [n],S 2 [n],…,S d [n],…,S D [n]],S d [n] represents the level state of the d-phase bridge arm of the traction converter in the [n]th sampling period, where the level state of the bridge arm is composed of a number of discrete values ​​and is determined by the topological structure of the traction converter, and D is the total number of bridge arms of the traction converter; u[n] is the input vector of the system in the [n]th sampling period, where the representation of u[n] satisfies u[n]=[u1[n],u2[n],…,u q [n],…,u Q [n]] T ,u q [n] is the qth input variable, T represents the transpose of the matrix, Q is the total number of input variables; C is the time-varying parameter matrix composed of non-vector terms in the state space equation related to the control quantity of the system electrical performance; S12: Construct a prediction model for the thermal performance control quantity of the traction converter to be tested, which is expressed as: In the formula, is the predicted total energy consumption of the jth power device in the kth power module of the traction converter in the [n+1]th sampling period, k = 1, 2, ..., K, K is the total number of power modules in the traction converter, j = 1, 2, ..., J k , J k is the total number of power devices in the kth power module; S k [n] represents the level state in the kth power module of the traction converter in the [nth]th sampling period; i k [n+1] is the predicted value of the current flowing through the kth power module of the traction converter in the [n+1]th sampling period, T kj [n] is the junction temperature of the jth power device in the kth power module in the [n]th sampling period; and are the conduction loss and switching loss of the jth power device in the kth power module of the traction converter, respectively. Their values ​​are calculated based on the junction temperature T kj [n] and the current prediction value i k [n+1] determined; and They are functions for determining whether the j-th power device in the k-th power module generates conduction loss and switching loss, respectively.

2. The method for dynamic optimization control of thermal performance of traction converter according to claim 1, characterized in that: The step of adjusting the electric and thermal performance of the traction converter to be tested according to the target system level state combination includes: When the traction converter to be tested is in a normal operating state or a fault state, the overall temperature of the traction converter to be tested is reduced, and the temperature differences of the bridge arms tend to be consistent, or; Reduce the average temperature of a single power module to make the temperature differences between the power devices in the module more consistent; The fault state is used to indicate that an open circuit fault occurs in a power device or a clamping diode in a power module of the traction converter.

3. The method for dynamic optimization control of thermal performance of traction converter according to claim 1, characterized in that: The electrical performance control quantity of the traction converter includes: phase current, DC voltage, average power, or transient speed; the thermal performance control quantity of the traction converter includes device junction temperature or device energy consumption.

4. The method for dynamic optimization control of thermal performance of traction converter according to claim 1, characterized in that: The S2 includes: S21: constructing an electrical performance control objective function according to the prediction model of the electrical performance control quantity, expressed as: In the formula, g electric (l)[n] is the value of the control objective function of the system electrical performance under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [nth]th sampling period, where l = 1, 2, ..., L, L is the total number of all available combinations of the level states of each phase bridge arm of the traction converter; x p_ref [n] is the reference value of the pth state variable in the system electrical performance control quantity in the [nth]th system sampling period; x p [n+1](S(l)[n]) is the predicted value of the pth state variable in the system electrical performance control quantity in the [n+1]th sampling period under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [n]th sampling period, is the weight coefficient of the pth state variable in the system electrical performance control quantity in the [n]th sampling period; S22: Constructing a thermal performance control objective function based on the dynamic weighting of the standard deviation coefficient according to the prediction model of the thermal performance control amount, expressed as: In the formula, g thermal (l)[n] is the value of the system thermal performance control objective function under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [nth]th system sampling period, Var k () is the thermal performance control objective function of the kth power module of the traction converter based on the energy consumption variance, is the switching cycle t of the jth power device in the kth power module of the traction converter in the [n]th system sampling cycle before the [n]th system sampling cycle sw Total energy consumption in s is the system sampling period, J k ′ is the total number of power devices that can work normally in the kth power module of the traction converter, is the weight coefficient in the control objective function of the thermal performance control quantity of the kth power module of the traction converter in the [n]th system sampling period, N thermal The sampling period window size of the historical data used to calculate the weight coefficient of the system thermal performance control quantity is determined by the heat dissipation time constant τ of the power module. thermal With the system sampling period t s The ratio of mod(n,N thermal ) is the dividend n, N thermal As the modulo operation of the divisor, T avg_k [n] is the average value of the sum of the junction temperatures of the normal power devices in the kth power module in the [nth]th system sampling period, flag kj [n] is the flag bit of the operating status of the jth power device in the kth power module in the [nth]th system sampling cycle. When the power device can work normally, flag kj [n] = 1; when a power device fails or is unable to work normally due to the faulty power device, the corresponding power device operating status flag is kj [n]=0.

5. The method for dynamic optimization control of thermal performance of traction converter according to claim 1, characterized in that: The S3 includes: Construct a reward function to represent the performance normalization, expressed as: Where R(S(l)) is the value of the reward function based on performance normalization under the action of the level state combination S(l) of each phase arm of the lth group of traction converters, α electric is the normalized control objective function G of the system electrical performance control quantity electric The weight coefficient, G electric (S(l)) is the value of the normalized control objective function of the system electrical performance control quantity under the action of the level state combination S(l) of each phase bridge arm of the lth group of traction converters, α thermal is the normalized control objective function G of the system thermal performance control quantity thermal The weight coefficient, G thermal (S(l)) is the value of the normalized control objective function of the system thermal performance control quantity under the action of the level state combination S(l) of each phase arm of the traction converter of the lth group, min{} represents the value of the control objective function corresponding to the minimum of the system electrical performance or thermal performance control objective function selected from all available combinations of the level states of each phase arm of the traction converter, g electric (l)[n] is the value of the control objective function of the system electrical performance under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [n]th sampling period, g thermal (l)[n] is the value of the system thermal performance control objective function under the action of the level state combination S(l)[n] of each phase bridge arm of the lth group of traction converters in the [n]th system sampling period.

6. The method for dynamic optimization control of thermal performance of traction converter according to claim 1, characterized in that: In S4, the system level state combination set is expressed as: In the formula, S(l normal ) is the number of all available traction converter bridge arm level state combinations under normal system operation, and its number is consistent with the number of all possible traction converter bridge arm level state combinations under normal system operation, where l normal =1,2,…,L normal , L normal is the total number of all available traction converter phase arm level state combinations under normal system operation, is the combination of the bridge arm level states of all available traction converters when the system has the mth type of fault, where: is the total number of all available combinations of bridge arm level states of each phase of the traction converter when the system has the mth type of fault, m = 1, 2, ..., M, M is the total number of all possible open circuit fault types of power devices and clamping diodes in the system.

7. The method for dynamic optimization control of thermal performance of traction converter according to claim 6, characterized in that: The S5 includes: S51: Taking the maximum reward function value as the optimization goal, a one-step optimized reward function is established, which is expressed as: Where R(S(H)) represents the value of the reward function corresponding to the maximum reward function based on performance normalization under the action of the Hth group of traction converter bridge arm level state combination S(H), H∈{1,2,...,l,...,L}; max{} represents the value of the reward function corresponding to the maximum reward function selected from all available combinations of the bridge arm level states of the traction converter; S52: The combination S(H) of the bridge arm level states of each phase of the Hth group of traction converters is used as the system control command output within [n] sampling cycles to control the on / off state of each power device, thereby determining the heat generated by the power devices in the power module and the current path and voltage in the traction converter, so as to achieve dynamic optimization control of the thermal performance of the system when it is in normal operation or fault state.

8. A traction converter thermal performance dynamic optimization control system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method described in any one of claims 1 to 7 are implemented.

Images