WO2022160723A1 - 牵引变流器热场控制方法及系统 - Google Patents

牵引变流器热场控制方法及系统 Download PDF

Info

Publication number
WO2022160723A1
WO2022160723A1 PCT/CN2021/117167 CN2021117167W WO2022160723A1 WO 2022160723 A1 WO2022160723 A1 WO 2022160723A1 CN 2021117167 W CN2021117167 W CN 2021117167W WO 2022160723 A1 WO2022160723 A1 WO 2022160723A1
Authority
WO
WIPO (PCT)
Prior art keywords
thermal field
control
value
traction converter
objective function
Prior art date
Application number
PCT/CN2021/117167
Other languages
English (en)
French (fr)
Inventor
杨超
彭涛
黄啸林
阳春华
桂卫华
谢斐然
陶宏伟
陈志文
樊欣宇
Original Assignee
中南大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中南大学 filed Critical 中南大学
Publication of WO2022160723A1 publication Critical patent/WO2022160723A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
    • H02M5/00—Conversion of ac power input into ac power output, e.g. for change of voltage, for change of frequency, for change of number of phases
    • H02M5/40—Conversion of ac power input into ac power output, e.g. for change of voltage, for change of frequency, for change of number of phases with intermediate conversion into dc
    • H02M5/42—Conversion of ac power input into ac power output, e.g. for change of voltage, for change of frequency, for change of number of phases with intermediate conversion into dc by static converters
    • H02M5/44—Conversion of ac power input into ac power output, e.g. for change of voltage, for change of frequency, for change of number of phases with intermediate conversion into dc by static converters using discharge tubes or semiconductor devices to convert the intermediate dc into ac
    • H02M5/453—Conversion of ac power input into ac power output, e.g. for change of voltage, for change of frequency, for change of number of phases with intermediate conversion into dc by static converters using discharge tubes or semiconductor devices to convert the intermediate dc into ac using devices of a triode or transistor type requiring continuous application of a control signal
    • H02M5/458—Conversion of ac power input into ac power output, e.g. for change of voltage, for change of frequency, for change of number of phases with intermediate conversion into dc by static converters using discharge tubes or semiconductor devices to convert the intermediate dc into ac using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only

Definitions

  • the invention relates to the technical field of power electronics, in particular to a method and system for controlling a thermal field of a traction converter.
  • the traction converter As the core equipment of high-end rail transit equipment such as permanent magnet high-speed trains, the traction converter is responsible for the energy supply and conversion of the traction drive of the train. As a high-incidence source of train failures, once the failure of the traction converter will affect the normal operation of the train, in severe cases, the train will be forced to stop. It can be seen that its reliability directly affects the safe operation level of the whole vehicle. According to the statistical results of various rail transit equipment operating agencies, the power module is the main source of the failure of the traction converter, and it is one of the high-frequency maintenance components of the train, which brings high maintenance costs.
  • thermal cycle stress accumulation and shock are the main reasons for the failure of power modules (such as IGBT (Insulated Gate Bipolar Translator, insulated gate bipolar transistor) modules). Therefore, reducing the thermal cycle stress intensity of the power module during the working process has become the most effective solution to improve the operational reliability of the power module and prolong its service life.
  • the active temperature management technology has received the most attention and has become the world's major rail transit scientific research. The main research direction of the institution. Then, the existing research mainly focuses on the improvement of thermal stress of a single power module, and the research on the improvement of thermal stress of a traction converter composed of multiple power modules is in its infancy.
  • the purpose of the present invention is to disclose a method and system for controlling the thermal field of the traction converter, so as to realize the uniform distribution of the thermal field of the traction converter by intelligently regulating the heat generated by each power device module in the traction converter, thereby effectively prolonging the traction converter.
  • the overall service life of the converter improves the reliability of train operation and reduces equipment maintenance costs.
  • the present invention discloses a method for controlling the thermal field of a traction converter, comprising:
  • Step S1 According to the relationship between the power consumption generated by each power module, the level state of the bridge arm to which it belongs, and the non-thermal field control quantity indirectly related to the power consumption, construct a power consumption prediction model of each power module corresponding to the system thermal field control quantity;
  • Step S2 constructing a thermal field distribution control objective function based on the power consumption variance according to the power consumption prediction model corresponding to the thermal field control quantity of the system, and then combining the thermal field distribution control objective function based on the power consumption variance with each of the non-thermal field control quantities.
  • the control objective functions are combined to construct a performance-based normalized control objective function;
  • Step S3 establishing an initial reward function based on performance normalization according to the normalized control objective function, and optimizing the initial reward function as a target reward function that dynamically adjusts the thermal weight coefficient according to the outer loop feedback value;
  • Step S4 in all possible combinations of the level states of the bridge arms of each phase of the traction converter in total L, select the system level state combination that maximizes the target reward function value as the system control command output to achieve the corresponding power. Intelligent regulation of module heat.
  • the thermal field distribution control objective function described in the step S2 is used to make the total power consumption of each power module of the traction converter tend to be the same within a period of time, thereby generating a similar amount of heat, and making the traction variable
  • the heat generated by each power module in the converter system will form a thermal system with uniform heat distribution; the thermal field refers to the thermal system formed by the heat generated by each power module due to its own power consumption during the operation of the traction converter .
  • the non-thermal field control quantity is current.
  • the step S1 specifically includes the following steps:
  • x m [n+1] A m [n]x m [n]+B m (S[n])u[n]+C[n]
  • x m [n+1] is the state vector for predicting the non-thermal field control variables of the mth type of system in the [n+1]th system sampling period
  • x m [n] is the [n]th system sampling period
  • the state vector of the non-thermal field control variables of the m-th type of system is the p-th state variable in the state space equation related to the non-thermal field control quantity of the m-th type of system,
  • the value is the sampling value of the mth type of non-thermal field control variable in the [n]th system sampling period
  • P is the total number of state variables in the state space equation related to the mth type of system non-thermal field control variable
  • m 1 ,2,...,M
  • M is the total number of non-thermal field control variables of the system
  • a m [n] is the time-varying parameter matrix related to state variables in the state space equation related to the non-thermal field control variables of the mth type of system
  • S12 Establish the relationship between the power consumption generated by the power module and the level state of the bridge arm to which it belongs, and build a power consumption-based prediction model of the system thermal field control quantity, the expression of which is:
  • S k [n] represents the level state of the k-th phase bridge arm of the traction converter in the [n]th system sampling period
  • i k [n+1] is the predicted value of the current flowing through the k-th phase bridge arm of the traction converter in the [n+1]th system sampling period
  • the value i k [n+1] is related to the junction temperature T kj [n] of the power module itself in the [n]th system sampling period; and are the functions of judging whether the j-th power module produces conduction loss and switching loss, respectively; wherein, i k [n+1] is the predicted value of the system current control amount obtained by S11.
  • the step S2 specifically includes:
  • g m [n] is the value of the control objective function of the non-thermal field control quantity of the mth type of system in the [n]th system sampling period
  • f m is the control objective function of the non-thermal field control quantity of the mth type of system
  • g el [n] is the value of the control objective function of the system thermal field control variable in the [n]th system sampling period
  • Var( ) is the thermal field distribution control objective function based on the power consumption variance
  • the control quantity type is the general name of the system non-thermal field control quantity and the system thermal field control quantity, namely type ⁇ m,el ⁇ ; is the sum of the performance-based control objective function values of the num_type state variable of a certain control variable type of the system under the combined action of the level state of each phase bridge arm of the lth group of traction converters; NUM_type is the content of a certain control variable type of the system
  • g m [n] or gel [n] is calculated according to step S21 or S22 as the performance-based control objective function of the num_type th state variable value ).
  • the normalized expression G type (S(l)) is 1, it means that under the combined action of the level states of the bridge arms of each phase of the l-th group of traction converters, a certain control variable type of the system
  • the control objective function can reach the minimum value, that is, the performance of the control variable is optimal at this time; if the normalized expression G type (S(l)) is less than 1, the value of G type (S(l)) Represents the proportion when the performance of a certain control variable type of the system is its optimal performance under the combined action of the level states of each phase bridge arm of the lth traction converter, expressed as a percentage.
  • the step S3 specifically includes the following steps:
  • R(S(l)) is the value of the initial reward function based on performance normalization under the action of the level state combination S(l) of each phase bridge arm of the l-th traction converter
  • ⁇ m is the m-th class
  • the weight coefficient of the normalized control objective function G m of the non-thermal field control variables of the system, and the value range of ⁇ m is determined according to the actual application
  • G m (S(l)) is the bridge arm of each phase of the lth group of traction converters
  • ⁇ el is the weight coefficient of the normalized control objective function G el of the thermal field control variable, and
  • the value range of ⁇ el needs to be determined according to the actual application
  • G el (S(l)) is the power consumption variance of the thermal field control amount under the action of the level state combination S(
  • ⁇ el_dy is the value of dynamically adjusting the thermal weight coefficient based on the feedback value of the outer loop
  • f el ( ) is a function describing the relationship between the feedback value of the outer loop and the thermal weight coefficient in the reward function
  • It is the value of the thermal weight coefficient set by the system/user before the value of the control variable of the system control outer loop changes, is the value of the thermal weight coefficient set by the system/user after the value of the system control outer loop control variable changes and reaches stability
  • ⁇ start is the initial value before the value of the system control outer loop control variable changes
  • ⁇ ref is The target reference value that the system is expected to achieve after the value of the system control outer loop control variable changes
  • ⁇ [n] is the sampling value of the system control outer loop control variable in the [n]th system sampling period
  • control outer loop feedback value includes the sensor sampling value of the outer loop control variable in the system closed-loop control and the system reference given/user set value.
  • the step S4 specifically includes the following steps:
  • S42 Output the level state combination S(H) of each phase bridge arm of the H-th traction converter as a system control command within [n] system sampling periods to control the on-off state of each power device, thereby realizing the power module Intelligent regulation of heat.
  • the present invention also discloses a traction converter thermal field control system, comprising a memory, a processor and a computer program stored in the memory and running on the processor, the processor executing the computer program When implementing the corresponding steps of the above method.
  • the uniform distribution of the thermal field of the traction converter is realized.
  • the method is easy to implement, does not require additional hardware equipment, and is of great significance for extending the overall service life of the traction converter, improving the reliability of train operation, and reducing equipment maintenance costs.
  • FIG. 1 is a topology diagram of a three-level traction inverter according to an embodiment of the present invention.
  • FIG. 2 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.
  • FIG. 3 is a flow chart of an intelligent life extension control method for a traction converter with uniform thermal field distribution according to an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of temperature fluctuations of four power modules of a U-phase bridge arm before and after an intelligent life extension control strategy is adopted in an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of temperature fluctuations of the four power modules of the V-phase bridge arm under the intelligent life extension control strategy in the dynamic process of the permanent magnet traction motor mechanical angular velocity from 352rad/s to constant velocity in the embodiment of the present invention.
  • the present invention is also referred to as "a traction converter thermal field control method”: “a traction converter intelligent life extension control method with uniform thermal field distribution”.
  • the follow-up system is the same and will not be described in detail.
  • this embodiment refers to a three-level traction converter system of a certain type of permanent magnet 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 structure As shown in Figure 1.
  • This example will take the intelligent life extension control of the three-phase three-level inverter as an example to illustrate.
  • the traction three-level inverter system adopts the speed-current double closed-loop control structure, and the control outer loop is the speed loop.
  • the given value of the system electromagnetic torque can be obtained; then, the full-speed permanent magnet motor control strategy of the maximum torque-to-current ratio (MTPA) and the maximum torque-to-voltage ratio (MTPV) mixture is used to obtain the system dq-axis current.
  • the given value is used as the given input of the current inner loop control strategy; finally, the system level state combination that maximizes the reward function value is selected by the intelligent life extension control strategy as the system control command output to realize the system dq axis current supply.
  • the tracking of the fixed value and the intelligent regulation of the heat of the power module, the control principle block diagram is shown in Figure 2,
  • an intelligent life extension control method for a traction converter with a uniform thermal field distribution includes the following steps.
  • Step S1 According to the relationship between the power consumption generated by each power module, the level state of the bridge arm to which it belongs, and the non-thermal field control quantity indirectly related to the power consumption, construct a power consumption prediction model of each power module corresponding to the system thermal field control quantity. In other words, the relationship between each control variable of the system and the level state of the traction converter is established, and the prediction model of each control variable of the system is constructed.
  • the intelligent life extension control of the traction inverter in this embodiment is in the current inner loop control link in the traction converter speed-current double closed-loop control structure. Therefore, in this embodiment, the The system control quantities are current and thermal field (temperature).
  • the thermal field refers to the thermal system formed by the heat generated by each power module due to its own power consumption during the operation of the traction converter.
  • Step 11 Establish the relationship between the non-thermal field control quantity of the system and the level state of the bridge arm to which it belongs, and build a prediction model for the non-thermal field control quantity of the system, which can be expressed as:
  • x m [n+1] A m [n]x m [n]+B m (S[n])u[n]+C[n] (1)
  • x m [n+1] is the state vector for predicting the non-thermal field control variables of the mth type of system in the [n+1]th system sampling period
  • x m [n] is the [n]th system sampling period
  • the state vector of the non-thermal field control variables of the m-th type of system is the p-th state variable in the state space equation related to the non-thermal field control quantity of the m-th type of system, and its value is the sampling value of the m-th type of non-thermal field control quantity in the [n]th system sampling period
  • a m [n] is the The time-varying parameter matrix related to the state variables in the state space equation related to the non-thermal field control variables of the mth type of system, B m
  • u q is the qth input variable
  • Q is the total number of input variables
  • C m is the time-varying parameter matrix of other terms in the state space equation related to the non-thermal field control variables of the mth type of system.
  • sk1 , sk2 , sk3 and sk4 respectively represent the control signals that determine the on-off state of the four power modules of each bridge arm, “1” means the control power module is in the on state, “0” means the control power module In the off state, the corresponding relationship is shown in Figure 1.
  • the non-thermal field control amount of the system is only the current amount.
  • ⁇ e [n] is the electrical angle of the system in the [n]th system sampling period, which can be calculated by multiplying the mechanical angle ⁇ m [n] measured by the system speed sensor by the number of pole pairs n p of the permanent magnet traction motor obtained, as shown in Figure 2.
  • Step 12 Establish the relationship between the power consumption generated by the power module and the level state of the bridge arm to which it belongs, and build a power consumption-based prediction model of the system thermal field control quantity, the expression of which is:
  • i k [n+1] can be predicted and calculated by the formula (1)
  • the d-axis stator current of the permanent magnet traction motor in the [n+1]th system sampling period in the dq coordinate system id [n+1] and q-axis stator current i q [n+1] are calculated by inverse Park transform and Clark inverse transform.
  • the formula can be expressed as:
  • the expression on the right side of the equation (2) is and
  • the expression of can be a fitting function, and the fitted data comes from the user data manual of the power module manufacturer.
  • Step S2 constructing a thermal field distribution control objective function based on the power consumption variance according to the power consumption prediction model corresponding to the thermal field control quantity of the system, and then combining the thermal field distribution control objective function based on the power consumption variance with each of the non-thermal field control quantities.
  • the control objective functions are combined to construct a performance-based normalized control objective function. In other words, namely: establish a thermal field distribution control objective function based on power consumption variance, and construct a performance-based normalized control objective function. Specifically include:
  • Step 21 Establish the control objective function of the non-thermal field control variables of the system, which can be expressed as:
  • g m [n] is the value of the control objective function of the non-thermal field control quantity of the mth type of system in the [n]th system sampling period
  • f m is the control objective function of the non-thermal field control quantity of the mth type of system , which can be designed according to the needs of the control target.
  • there are expressions such as the square of the difference, the absolute value, and the 2-norm; is a vector consisting of the system reference given value/user set value of the non-thermal field control variables of the mth type of system in the [n]th system sampling period, is the reference value of the p-th state variable in the non-thermal field control variables of the m-th type of system.
  • control objective function of the non-thermal field control quantity is:
  • Step 22 Establish a thermal field distribution control objective function based on power consumption variance, which can be expressed as:
  • g el [n] is the value of the control objective function of the system thermal field control variable in the [n]th system sampling period
  • Var( ) is the thermal field distribution control objective function based on the power consumption variance
  • ts is the system sampling period cycle.
  • Step 23 Construct a performance-based normalized control objective function whose unified expression is:
  • the control quantity type is the general name of the system non-thermal field control quantity and the system thermal field control quantity, namely type ⁇ m,el ⁇ ; is the sum of the performance-based control objective function values of the num_type state variable of a certain control variable type of the system under the combined action of the level state of each phase bridge arm of the lth group of traction converters, and NUM_type is the content of a certain control variable type of the system
  • g m [n] or gel [n] is calculated according to step S21 or S22 as the performance-based control objective function of the num_type th state variable value ).
  • the normalized control objective function of the current control quantity is:
  • the normalized control objective function of the thermal field control quantity is:
  • Step S3 establishing an initial reward function based on performance normalization according to the normalized control objective function, and optimizing the initial reward function as an objective reward function that dynamically adjusts the thermal weight coefficient according to the outer loop feedback value.
  • the relationship between the feedback value of the control outer loop and the thermal weight coefficient is established, and a reward function that is dynamically adjusted based on the thermal weight coefficient is constructed.
  • control outer loop of the three-level traction converter is a speed loop.
  • outer-loop control quantity in the closed-loop control of the three-level traction converter system is the mechanical angular velocity ⁇ m of the permanent magnet traction motor, as shown in FIG. 2 .
  • Step 31 Establish an initial reward function based on performance normalization, which can be expressed as:
  • R(S(l)) is the value of the initial reward function based on performance normalization under the action of the level state combination S(l) of each phase bridge arm of the l-th traction converter, and ⁇ m is the m-th class
  • G m the non-thermal field control variables of the system.
  • the value of the normalized control objective function of the non-thermal field control variable of the mth type system under the action of the level state combination S(l); ⁇ el is the weight coefficient of the normalized control objective function G el of the thermal field control variable, which is The value range is determined according to the actual application, G el (S(l)) is the power consumption variance based on the thermal field control amount under the action of the level state combination S(l) of each phase bridge arm of the l-th traction converter.
  • the thermal field distribution normalization controls the value of the objective function.
  • Step 32 Establish the relationship between the outer loop feedback value and the thermal weight coefficient in the reward function, which can be expressed as:
  • ⁇ el_dy is the value of dynamically adjusting the thermal weight coefficient based on the feedback value of the outer loop
  • f el ( ) is a function describing the relationship between the feedback value of the outer loop and the thermal weight coefficient in the reward function
  • It is the value of the thermal weight coefficient set by the system/user before the value of the control variable of the system control outer loop changes, is the value of the thermal weight coefficient set by the system/user after the value of the system control outer loop control variable changes and reaches stability
  • ⁇ start is the initial value before the value of the system control outer loop control variable changes
  • ⁇ ref is The target reference value that the system is expected to achieve after the value of the control variables of the system control outer loop changes, which can be obtained from the system given/user setting
  • ⁇ [n] is the system control outer loop control in the [n]th system sampling period The sampled value of the variable.
  • control variable ⁇ of the system control outer loop is the mechanical angular velocity ⁇ m of the permanent magnet traction motor.
  • the functional expression of the outer loop feedback value and the thermal weight coefficient in the reward function is:
  • ⁇ start 352 rad/s
  • ⁇ ref 530 rad/s.
  • Step 33 Construct a target reward function dynamically adjusted based on the thermal weight coefficient, which can be expressed as:
  • the reward function dynamically adjusted based on the thermal weight coefficient is:
  • Step S4 in all possible combinations of the level states of the bridge arms of each phase of the traction converter in total L, select the system level state combination that maximizes the target reward function value as the system control command output to achieve the corresponding power. Intelligent regulation of module heat.
  • the specific implementation may include the following steps.
  • Step 41 Taking the maximum value of the reward function dynamically adjusted based on the thermal weight coefficient as the optimization goal, establish a calculation function for one-step optimization, which can be expressed as:
  • Step 42 Output the level state combination S(H) of each phase bridge arm of the H-th traction converter as a system control command within [n] system sampling periods to control the on-off state of each power device, thereby realizing power Intelligent regulation of module heat.
  • i 1 [n], i 2 [n], i 3 [n], ⁇ e [n], and ⁇ m [n] are obtained through system sensor sampling; further, through the system outer loop control id_ref [n] and iq_ref [n] can be obtained by the policy; in addition, ⁇ ref and ⁇ start can be obtained through the system reference command/user setting.
  • the temperature changes of the U-phase four power modules are shown in Figure 4. It can be seen that the thermal field distribution of the traction converter system using the intelligent life extension control strategy is more uniform, and the temperature of each power module of the bridge arm tends to be consistent. Compared with the traction converter system that does not adopt the intelligent life extension control strategy, the control method of this patent can make the temperature and fluctuation of each power module similar. According to the relevant research results, the life consumption of each power module will also be reduced. tend to be similar, the overall service life of the traction converter will effectively avoid the "barrel short board" effect, thereby achieving the extension of its overall service life.
  • the thermal weight coefficient will continue to increase; further, in the reward function based on the thermal weight coefficient, the thermal distribution control
  • the value of the normalized control objective function corresponding to the amount of ⁇ H also increases accordingly; as a result, in the one-step optimization, under the action of the level state combination S(H) of each phase bridge arm of the H group of traction converters, the system heat distribution
  • the situation will be gradually improved, its thermal field distribution will gradually tend to be evenly divided, and the temperature and its fluctuation of each power module will also gradually tend to be consistent;
  • the intelligent regulation of the temperature of the power module thereby achieving the goal of extending the overall service life of the traction converter system.
  • the temperature conditions of the four power modules of the V phase during the acceleration process are shown in Figure 5.
  • the present embodiment provides an intelligent life extension control system for a traction converter with a uniform thermal field distribution, including a memory, a processor, and a computer program stored in the memory and running on the processor, The steps of the above-described methods are implemented when the processor executes the computer program.
  • the methods and systems disclosed by the above two embodiments of the present invention respectively realize uniform distribution of the thermal field of the traction converter by intelligently regulating the heat generated by each power device module in the traction converter.
  • the method is easy to implement, does not require additional hardware equipment, and is of great significance for extending the overall service life of the traction converter, improving the reliability of train operation, and reducing equipment maintenance costs.

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Inverter Devices (AREA)
  • Electric Propulsion And Braking For Vehicles (AREA)

Abstract

本发明涉及电力电子技术领域，公开一种牵引变流器热场控制方法，以实现牵引变流器热场的均匀分布。方法包括：构建各功率模块对应系统热场控制量的功耗预测模型；根据对应系统热场控制量的功耗预测模型构建基于功耗方差的热场分布控制目标函数，然后将基于功耗方差的热场分布控制目标函数与各非热场控制量的控制目标函数相结合以构建基于性能的归一化控制目标函数；根据归一化控制目标函数建立基于性能归一化的初始奖励函数，并将初始奖励函数优化为根据外环反馈值动态调整热权重系数的目标奖励函数；选取使得目标奖励函数值最大化的系统电平状态组合作为系统控制指令输出，实现相对应功率模块热量的智能调控。

Description

牵引变流器热场控制方法及系统 技术领域
本发明涉及电力电子技术领域，尤其涉及一种牵引变流器热场控制方法及系统。
背景技术
随着我国高铁事业的快速发展，运行时速高、运行里程长等高铁发展趋势已向高安全、高可靠运行的高铁发展新诉求发生转变。特别是，以主动安全保障为核心的高铁安全可靠运行技术成为当前高速列车科技发展的趋势，国家“十三五”重点研发计划先进轨道交通重点专项更是将此列为我国轨道交通需要优先发展的方向。
牵引变流器作为永磁高速列车等高端轨道交通装备的核心设备，负责列车牵引驱动的能量供给与转换。作为列车的高发故障源，牵引变流器一旦发生故障将影响列车的正常运行，严重时将导致列车被迫停车，由此可见，其可靠性直接影响着整车的安全运行水平。据各轨道交通装备运营机构的统计结果表明，功率模块是牵引变流器故障的主要来源，属于列车高频次维修器件之一，带来较高维修成本。
据实验结果表明，除遭受过电应力和其他极端异常情况外，热循环应力累积和冲击是功率模块(如：IGBT(Insulated Gate Bipolar Translator，绝缘栅双极型晶体管)模块)的失效主因。因此，降低工作过程中功率模块的热循环应力强度，成为提高功率模块运行可靠性、延长使用寿命的最为有效方案，其中，又以主动温度管理技术最受关注，已成为世界各大轨道交通科研机构的重点研究方向。然后，现有研究主要针对单个功率模块的热应力改善，对由多个功率模块组成的牵引变流器热应力改善的研究处于起步阶段。
目前，在牵引变流器集成化、模块化的普及和应用下，一旦某个功率模块因热应力损坏，整个牵引变流器设备将被替换维修，其中包括大量还有较长剩余寿命的其它正常功率模块。事实上，当前列车牵引变流器都以电气性能为主的控制目标，且因功率模块的热应力难以简单的加入到传统的闭环控制中，造成牵引变流器的热应力状况被忽视，特别是在多电平牵引变流器的应用中，牵引变流器中各功率模块的热应力不平衡问题更为凸显。这样不仅会造成资源浪费，维修成本的提升，且将降低牵引变流器整体服役寿命和运行可靠性。然而，仅关注单个功率器件模块的热应力改善/主动温度管理技术，将难以实现对整个牵引变流器热应力的改善。
因此，如何实现牵引变流器各功率器件的主动温度/热应力平衡成为一个亟待解决的关键 技术，此问题的解决将有利于提高牵引变流器乃至整车的整体服役寿命和运行可靠性，降低资源浪费和维修成本，具有重要的意义。
发明内容
本发明目的在于公开一种引变流器热场控制方法及系统，以通过智能调控牵引变流器中各功率器件模块产生的热量，实现牵引变流器热场的均匀分布，进而有效延长牵引变流器整体服役寿命，提高列车运行可靠性水平，降低设备维护成本。
为达上述目的，本发明公开一种牵引变流器热场控制方法，包括：
步骤S1：根据各功率模块产生功耗与所属桥臂电平状态和与功耗间接关联的非热场控制量的关系，构建各功率模块对应系统热场控制量的功耗预测模型；
步骤S2：根据对应系统热场控制量的功耗预测模型构建基于功耗方差的热场分布控制目标函数，然后将基于功耗方差的热场分布控制目标函数与各所述非热场控制量的控制目标函数相结合以构建基于性能的归一化控制目标函数；
步骤S3：根据所述归一化控制目标函数建立基于性能归一化的初始奖励函数，并将所述初始奖励函数优化为根据外环反馈值动态调整热权重系数的目标奖励函数；
步骤S4：在总数为L的牵引变流器各相桥臂电平状态所有可能组合中，选取使得所述目标奖励函数值最大化的系统电平状态组合作为系统控制指令输出，实现相对应功率模块热量的智能调控。
优选地，所述步骤S2所述的热场分布控制目标函数用于在一段时间内，使得牵引变流器各功率模块总功耗趋于相同，进而产生趋于相近的热量，并使得牵引变流器系统中各功率模块所产生的热量将形成热量分布均匀的一种热系统；所述热场是指牵引变流器运行中各功率模块因自身功耗而产生的热量所形成的热系统。
优选地，所述非热场控制量为电流。
优选地，所述步骤S1具体包括以下步骤：
S11：建立系统非热场控制量与所属桥臂电平状态的关系，构建系统非热场控制量的预测模型，表示为：
x m[n+1]＝A m[n]x m[n]+B m(S[n])u[n]+C[n]
式中，x m[n+1]为预测第[n+1]个系统采样周期内第m类系统非热场控制量的状态向量，x m[n]为第[n]个系统采样周期第m类系统非热场控制量的状态向量，
Figure PCTCN2021117167-appb-000001
Figure PCTCN2021117167-appb-000002
为与第m类系统非热场控制量有关的状态空间方程中的第p个状态变量，
Figure PCTCN2021117167-appb-000003
数值为第m类非热场控制量在第[n]个系统采样周期内的采样值，P为与第m类系统非热场控制量有关的状态空间方程中的状态变量总数，m＝1,2,…,M，M为系统非热场控制量的总数；A m[n]为与第m类系统非热场控制量有关的状态空间方程中与状态变量有关的时变参数矩阵，B m(S[n])为与第m类系统非热场控制量有关的状态空间方程中的输入变量的时变参数矩阵，其表达式与S[n]有关，S[n]为第[n]个系统采样周期内牵引变流器各相桥臂的电平状态所有可能组合的集合，S＝[S 1,S 2,…,S k,…,S K]，S k表示牵引变流器第k相桥臂的电平状态，其中桥臂的电平状态由若干离散数值组成，具体数量及数值由牵引变流器的拓扑结构所决定，K为牵引变流器的桥臂总数；u[n]为输入向量，u＝[u 1,u 2,…,u q,…,u Q] T，u q为第q个输入变量，Q为输入变量总数；C m为与第m类系统非热场控制量有关的状态空间方程中其它项的时变参数矩阵；
S12：建立功率模块产生功耗与所属桥臂电平状态的关系，构建基于功耗的系统热场控制量的预测模型，其表达式为：
Figure PCTCN2021117167-appb-000004
式中，
Figure PCTCN2021117167-appb-000005
为牵引变流器第k相桥臂第j个功率模块在第[n+1]个系统采样周期内的总体功耗预测值，k＝1,2,…,K，j＝1,2,…,J，J为牵引变流器中桥臂功率模块的总数；S k[n]表示在第[n]个系统采样周期内牵引变流器第k相桥臂的电平状态；i k[n+1]为在第[n+1]个系统采样周期内流经牵引变流器第k相桥臂的电流预测值；
Figure PCTCN2021117167-appb-000006
和
Figure PCTCN2021117167-appb-000007
分别为牵引变流器第k相桥臂第j个功率模块的导通损耗和切换损耗，数值与第[n+1]个系统采样周期内流经牵引变流器第k相桥臂的电流值i k[n+1]和第[n]个系统采样周期内功率模块本身结温T kj[n]有关；
Figure PCTCN2021117167-appb-000008
和
Figure PCTCN2021117167-appb-000009
分别为判断第j个功率模块是否产生导通损耗和产生切换损耗的函数；其中，i k[n+1]根据S11得到的系统电流控制量的预测值。
优选地，所述步骤S2具体包括：
S21:建立系统非热场控制量的控制目标函数，表示为：
Figure PCTCN2021117167-appb-000010
式中，g m[n]为第[n]个系统采样周期内第m类系统非热场控制量的控制目标函数的值，f m为第m类系统非热场控制量的控制目标函数；
Figure PCTCN2021117167-appb-000011
为第[n]个系统采样周期内第m类系统非热场控制量的系统参考给定值/用户设定值组成的向量，
Figure PCTCN2021117167-appb-000012
Figure PCTCN2021117167-appb-000013
为第m类系统非热场控制量中第p个状态变量的参考值；
S22：建立基于功耗方差的热场分布控制目标函数，表示为：
Figure PCTCN2021117167-appb-000014
式中，g el[n]为第[n]个系统采样周期内系统热场控制量的控制目标函数的值，Var(·)为基于功耗方差的热场分布控制目标函数；
Figure PCTCN2021117167-appb-000015
为第[n]个系统采样周期内牵引变流器第k相桥臂第j个功率模块在[n]个系统采样周期前的一个开关周期t sw内的总功耗，t s为系统采样周期；
S23：构建基于性能的归一化控制目标函数，其统一的表达式为：
Figure PCTCN2021117167-appb-000016
式中，G type(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下系统某一控制量type基于性能的归一化控制目标函数的值，其中，G type(S(l))∈(0,1]，l＝1,2，…，L，L为牵引变流器各相桥臂电平状态所有可能组合的总数，系统某一控制量type为系统非热场控制量和系统热场控制量的统称，即type∈{m,el}；
Figure PCTCN2021117167-appb-000017
为第l组牵引变流器各相桥臂电平状态组合作用下系统某一控制量type第num_type个状态变量的基于性能的控制目标函数值之和；NUM_type为系统某一控制量type所含状态变量的总数；min{·}表示在总数为L的牵引变流器各相桥臂电平状态所有可能组合S＝[S 1,S 2,…,S k,…,S K]中选取使得控制目标函数最小时对应的控制目标函数的值；
其中，当type为系统非热场控制量或热场控制量时，根据步骤S21或S22计算得到g m[n]或g el[n]，作为第num_type个状态变量的基于性能的控制目标函数值
Figure PCTCN2021117167-appb-000018
)。
优选地，若所述的归一化表达式G type(S(l))为1时，表示在第l组牵引变流器各相桥臂电平状态组合作用下系统某一控制量type的控制目标函数能达到最小值，即此时该控制量的性能最优；若所述的归一化表达式G type(S(l))小于1时，G type(S(l))的数值表示在第l组牵引变流器各相桥臂电平状态组合作用下系统某一控制量type的性能为其性能最优时的占比，用百分比表示。
优选地，所述步骤S3具体包括以下步骤：
S31：建立基于性能归一化的初始奖励函数，表示为：
Figure PCTCN2021117167-appb-000019
式中，R(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下基于性能归一化的初始奖励函数的值，λ m为第m类系统非热场控制量的归一化控制目标函数G m的权重系数，且λ m取值范围结合实际应用确定，G m(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下第m类系统非热场控制量的归一化控制目标函数的值；λ el为热场控制量的归一化控制目标函数G el的权重系数，且λ el取值范围需结合实际应用确定，G el(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下热场控制量的基于功耗方差的热场分布归一化控制目标函数的值；
S32：建立外环反馈值与奖励函数中热权重系数的关系，表示为：
Figure PCTCN2021117167-appb-000020
式中，λ el_dy为基于外环反馈值的动态调整热权重系数的数值，f el(·)为描述外环反馈值与奖励函数中热权重系数关系的函数；
Figure PCTCN2021117167-appb-000021
为系统控制外环控制变量的数值发生变化前系统/用户所设定的热权重系数的数值，
Figure PCTCN2021117167-appb-000022
为系统控制外环控制变量的数值发生变化并到达稳定后系统/用户所设定的热权重系数的数值；θ start为系统控制外环控制变量的数值发生变化前的起始值，θ ref为系统控制外环控制变量的数值发生变化后系统预期达到的目标参考值，θ[n]为 第[n]个系统采样周期内系统控制外环控制变量的采样值；
S33：构建基于热权重系数动态调整的目标奖励函数R′(S(l))，表示为：
R′(S(l))＝∑λ mG m(S(l))+λ el_dyG el(S(l))。
优选地，所述控制外环反馈值包括系统闭环控制中外环控制变量的传感器采样值和系统参考给定/用户的设定值。
优选地，所述步骤S4具体包括以下步骤：
S41：以使基于热权重系数动态调整的奖励函数值最大为优化目标，建立一步优化的计算函数，表示为：
R′[H]＝max{R′(S(L))}
式中，R′[H]表示当第H组牵引变流器各相桥臂电平状态组合S(H)作用下奖励函数最大化所取到的数值，S(l)∈S＝[S 1,S 2,…,S k,…,S K]；max{·}表示在总数为L的牵引变流器各相桥臂电平状态所有可能组合S＝[S 1,S 2,…,S k,…,S K]中选取使得奖励函数最大化时对应的奖励函数的值。
S42：将第H组牵引变流器各相桥臂电平状态组合S(H)作为[n]个系统采样周期内的系统控制指令输出，控制各功率器件的导断状态，从而实现功率模块热量的智能调控。
为达上述目的，本发明还公开一种牵引变流器热场控制系统，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述计算机程序时实现上述方法相对应的步骤。
本发明具有以下有益效果：
通过智能调控牵引变流器中各功率器件模块产生的热量，实现牵引变流器热场的均匀分布。该方法易于实施，无需额外硬件设备，对延长牵引变流器整体服役寿命，提高列车运行可靠性水平，降低设备维护成本等具有重要意义。
下面将参照附图，对本发明作进一步详细的说明。
附图说明
构成本申请的一部分的附图用来提供对本发明的进一步理解，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的不当限定。在附图中：
图1是本发明实施例的三电平牵引逆变器的拓扑图。
图2是本发明实施例的永磁列车三电平牵引逆变器系统整体控制原理框图。
图3是本发明实施例的一种热场均匀分布的牵引变流器智能延寿控制方法流程图。
图4是本发明实施例采用智能延寿控制策略前后U相桥臂4个功率模块的温度波动示意图。
图5是本发明实施例中永磁牵引电机机械角速度从352rad/s到恒速的动态过程中，采用智能延寿控制策略下V相桥臂4个功率模块的温度波动示意图。
具体实施方式
以下结合附图对本发明的实施例进行详细说明，但是本发明可以由权利要求限定和覆盖的多种不同方式实施。
实施例1
本实施例将本发明“牵引变流器热场控制方法”另称为：“热场均匀分布的牵引变流器智能延寿控制方法”。后续系统同理，不做赘述。
具体地，本实施例参照某型永磁列车三电平牵引变流器系统，该牵引变流器可进一步分为单相三电平整流器和三相三电平逆变器，具体的拓扑结构如图1所示。本实例将以三相三电平逆变器的智能延寿控制为例进行说明，牵引三电平逆变器系统采用转速-电流双闭环控制结构，控制外环为转速环，依据转速反馈经查找表，可得出系统电磁转矩的给定值；接着，采用最大转矩电流比(MTPA)和最大转矩电压比(MTPV)混合的全速域永磁电机控制策略，得到系统dq轴电流的给定值，并将此作为电流内环控制策略的给定输入；最后，由智能延寿控制策略选取使得奖励函数值最大化的系统电平状态组合作为系统控制指令输出，实现系统dq轴电流给定值的跟踪和功率模块热量的智能调控，其控制原理框图如图2所示，
本实施例中，某型永磁列车三电平牵引逆变器系统的主要参数如表1所示。
表1永磁列车三电平牵引逆变器系统的主要参数
参数 数值
定子电阻R s 0.07Ω
定子d轴电感L d 0.0037H
定子q轴电感L q 0.0096H
永磁体磁链Ψ r 0.625Wb
极对数n p 4
给定直流电压 3600V
永磁电机额定功率 600kW
系统采样时间t s 40μs
功率模块开关周期t sw 600μs
如图3所示，一种热场均匀分布的牵引变流器智能延寿控制方法包括如下步骤。
步骤S1：根据各功率模块产生功耗与所属桥臂电平状态和与功耗间接关联的非热场控制量的关系，构建各功率模块对应系统热场控制量的功耗预测模型。换言之，即：建立系统各控制量与牵引变流器电平状态的关系，构建系统各控制量的预测模型。
需要说明的是，如图2所示，本实施例中牵引逆变器的智能延寿控制处于牵引变流器转速-电流双闭环控制结构中的电流内环控制环节，因此，本实施例中的系统控制量为电流和热场(温度)。其中，热场是指牵引变流器运行中各功率模块因自身功耗而产生的热量所形成的热系统。
步骤11：建立系统非热场控制量与所属桥臂电平状态的关系，构建系统非热场控制量的预测模型，可表示为：
x m[n+1]＝A m[n]x m[n]+B m(S[n])u[n]+C[n]     (1)
式中，x m[n+1]为预测第[n+1]个系统采样周期内第m类系统非热场控制量的状态向量，x m[n]为第[n]个系统采样周期第m类系统非热场控制量的状态向量，
Figure PCTCN2021117167-appb-000023
Figure PCTCN2021117167-appb-000024
为与第m类系统非热场控制量有关的状态空间方程中的第p个状态变量，其数值为第m类非热场控制量在第[n]个系统采样周期内的采样值，P为与第m类系统非热场控制量有关的状态空间方程中的状态变量总数，m＝1,2,…,M，M为系统非热场控制量的总数；A m[n]为与第m类系统非热场控制量有关的状态空间方程中与状态变量有关的时变参数矩阵，B m(S[n])为与第m类系统非热场控制量有关的状态空间方程中的输入变量的时变参数矩阵，其表达式与S[n]有关，S[n]为第[n]个系统采样周期内牵引变流器各相桥臂的电平状态所有可能组合的集合，S＝[S 1,S 2,…,S k,…,S K]，S k表示牵引变流器第k相桥臂的电平状态，其中桥臂的电平状态由若干离散数值组成，具体数量及数值由牵引变流器的拓扑结构所决定，K为牵引变流器的桥臂总数；u[n]为输入向量，u＝[u 1,u 2,…,u q,…,u Q] T，u q为第q个输入变量，Q为输入变量总数；C m为与第m类系统非热场控制量有关的状态空间方程中其它项的时变参数矩阵。
需要说明的是，本实施例中，牵引三电平逆变器共有三相桥臂，每相桥臂由四个功率模 块组成，如图1所示，分别为U相桥臂、V相桥臂和W相桥臂，正常情况下各桥臂共有三种电平状态。因此，本实施例中，K＝3，S 1、S 2和S 3分别对应U相桥臂、V相桥臂和W相桥臂的电平状态。具体地，在本实施例中，第k相桥臂的电平状态S k，可表达为：
Figure PCTCN2021117167-appb-000025
式中，s k1，s k2，s k3和s k4分别表示决定各桥臂四个功率模块通断状态的控制信号，“1”表示控制功率模块处于导通状态，“0”表示控制功率模块处于关断状态，其对应关系如图1所示。
需要说明的是，如图2所示，本实施例中系统非热场控制量仅为电流量。具体地，x＝[i d,i q] T，M＝1，P＝2，Q＝1，u＝[u cd]，式(1)中时变参数矩阵分别为：
Figure PCTCN2021117167-appb-000026
Figure PCTCN2021117167-appb-000027
Figure PCTCN2021117167-appb-000028
式中，θ e[n]为第[n]个系统采样周期内系统的电角度，可以系统转速传感器测量到的机械角度θ m[n]乘以永磁牵引电机的极对数n p计算得到，如图2所示。
步骤12：建立功率模块产生功耗与所属桥臂电平状态的关系，构建基于功耗的系统热场控制量的预测模型，其表达式为：
Figure PCTCN2021117167-appb-000029
式中，
Figure PCTCN2021117167-appb-000030
为牵引变流器第k相桥臂第j个功率模块在第[n+1]个系统采样周期内的总体功耗预测值，k＝1,2,…,K，j＝1,2,…,J，J为牵引变流器中桥臂功率模块的总数；S k[n] 表示在第[n]个系统采样周期内牵引变流器第k相桥臂的电平状态；i k[n+1]为在第[n+1]个系统采样周期内流经牵引变流器第k相桥臂的电流预测值，且i k[n+1]具体根据步骤S11得到的系统电流控制量的预测值；
Figure PCTCN2021117167-appb-000031
和
Figure PCTCN2021117167-appb-000032
分别为牵引变流器第k相桥臂第j个功率模块的导通损耗和切换损耗，其数值与第[n+1]个系统采样周期内流经牵引变流器第k相桥臂的电流值i k[n+1]和第[n]个系统采样周期内功率模块本身结温T kj[n]有关；
Figure PCTCN2021117167-appb-000033
和
Figure PCTCN2021117167-appb-000034
分别为判断第j个功率模块是否产生导通损耗和产生切换损耗的函数。
需要注意的是，本实施例中，i k[n+1]可由式(1)预测计算得到的第[n+1]个系统采样周期内dq坐标系下永磁牵引电机的d轴定子电流i d[n+1]和q轴定子电流i q[n+1]，通过Park逆变换和Clark逆变换计算得到，具体地，公式可表达为：
Figure PCTCN2021117167-appb-000035
需要注意的是，本实施例中，式(2)中等式右边的表达式
Figure PCTCN2021117167-appb-000036
和
Figure PCTCN2021117167-appb-000037
的表达式可为拟合函数，拟合的数据来自功率模块厂商用户数据手册。
需要注意的是，本实施例中，判断是否产生导通损耗和产生切换损耗的函数
Figure PCTCN2021117167-appb-000038
和
Figure PCTCN2021117167-appb-000039
的表达式可由授权国家发明专利[一种牵引变流器器件结温在线计算方法及系统,ZL 201810490961.3]中推导出。
步骤S2：根据对应系统热场控制量的功耗预测模型构建基于功耗方差的热场分布控制目标函数，然后将基于功耗方差的热场分布控制目标函数与各所述非热场控制量的控制目标函数相结合以构建基于性能的归一化控制目标函数。换言之，即：建立基于功耗方差的热场分布控制目标函数，构建基于性能的归一化控制目标函数。具体包括：
步骤21：建立系统非热场控制量的控制目标函数，可表示为：
Figure PCTCN2021117167-appb-000040
式中，g m[n]为第[n]个系统采样周期内第m类系统非热场控制量的控制目标函数的值，f m为第m类系统非热场控制量的控制目标函数，可根据控制目标的需要进行设计，一般有差 的平方、绝对值、2-范数等表达方式；
Figure PCTCN2021117167-appb-000041
为第[n]个系统采样周期内第m类系统非热场控制量的系统参考给定值/用户设定值组成的向量，
Figure PCTCN2021117167-appb-000042
Figure PCTCN2021117167-appb-000043
为第m类系统非热场控制量中第p个状态变量的参考值。
需要说明的是，本实施例中，非热场控制量的控制目标函数为：
Figure PCTCN2021117167-appb-000044
式中，
Figure PCTCN2021117167-appb-000045
为第[n]个系统采样周期内dq坐标系下永磁牵引电机d轴定子电流和q轴定子电流的给定值，由系统控制外环计算后给出，具体如图2所示。
步骤22：建立基于功耗方差的热场分布控制目标函数，可表示为：
Figure PCTCN2021117167-appb-000046
式中，g el[n]为第[n]个系统采样周期内系统热场控制量的控制目标函数的值，Var(·)为基于功耗方差的热场分布控制目标函数；
Figure PCTCN2021117167-appb-000047
为第[n]个系统采样周期内牵引变流器第k相桥臂第j个功率模块在[n]个系统采样周期前的一个开关周期t sw内的总功耗，t s为系统采样周期。
需要注意的是，本实施例中，K＝3，J＝4。
需要注意的是，本实施例中，t sw＝600μs，t s＝40μs。
步骤23：构建基于性能的归一化控制目标函数，其统一的表达式为：
Figure PCTCN2021117167-appb-000048
式中，G type(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下系统某一控制量type基于性能的归一化控制目标函数的值，其中，G type(S(l))∈(0,1]，l＝1,2，…，L，L为牵引变流器各相桥臂电平状态所有可能组合的总数，系统某一控制量type为系统非热场控制量和系统热场控制量的统称，即type∈{m,el}；
Figure PCTCN2021117167-appb-000049
为第l组牵引变流器各相桥臂电平状态组合作用下系统某一控制量type第num_type个状态变量的基于性能的控制目标 函数值之和，NUM_type为系统某一控制量type所含状态变量的总数；min{·}表示在总数为L的牵引变流器各相桥臂电平状态所有可能组合S＝[S 1,S 2,…,S k,…,S K]中选取使得控制目标函数最小化时对应的控制目标函数的值。其中，当type为系统非热场控制量或热场控制量时，根据步骤S21或S22计算得到g m[n]或g el[n]，作为第num_type个状态变量的基于性能的控制目标函数值
Figure PCTCN2021117167-appb-000050
)。
需要注意的是，本实施例中，三电平牵引变流器各相桥臂电平状态所有可能组合的总数L＝27。
需要注意的是，本实施例中，电流控制量的归一化控制目标函数为：
Figure PCTCN2021117167-appb-000051
式中，P＝2，x 1[n+1]与S(l)的关系式由式(1)可得。
需要注意的是，本实施例中，热场控制量的归一化控制目标函数为：
Figure PCTCN2021117167-appb-000052
Figure PCTCN2021117167-appb-000053
式中，
Figure PCTCN2021117167-appb-000054
与S(l)的关系式由式(2)可得。
步骤S3：根据所述归一化控制目标函数建立基于性能归一化的初始奖励函数，并将所述初始奖励函数优化为根据外环反馈值动态调整热权重系数的目标奖励函数。换言之，即：建立控制外环反馈值与热权重系数的关系，构建基于热权重系数动态调整的奖励函数。
需要注意的是，本实施例中，三电平牵引变流器的控制外环为速度环。具体地，三电平牵引变流器系统闭环控制中的外环控制量为永磁牵引电机的机械角速度ω m，如图2所示。
步骤31：建立基于性能归一化的初始奖励函数，可表示为：
Figure PCTCN2021117167-appb-000055
式中，R(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下基于性能归一化的初始奖励函数的值，λ m为第m类系统非热场控制量的归一化控制目标函数G m的权重系数，其取值范围由需结合实际应用确定，G m(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下第m类系统非热场控制量的归一化控制目标函数的值；λ el为热场控制量的归一化控制目标函数G el的权重系数，其取值范围由需结合实际应用确定，G el(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下热场控制量的基于功耗方差的热场分布归一化控制目标函数的值。
需要注意的是，本实施例中，非热场权重系数λ m＝λ 1＝1。
步骤32：建立外环反馈值与奖励函数中热权重系数的关系，可表示为：
Figure PCTCN2021117167-appb-000056
式中，λ el_dy为基于外环反馈值的动态调整热权重系数的数值，f el(·)为描述外环反馈值与奖励函数中热权重系数关系的函数；
Figure PCTCN2021117167-appb-000057
为系统控制外环控制变量的数值发生变化前系统/用户所设定的热权重系数的数值，
Figure PCTCN2021117167-appb-000058
为系统控制外环控制变量的数值发生变化并到达稳定后系统/用户所设定的热权重系数的数值；θ start为系统控制外环控制变量的数值发生变化前的起始值，θ ref为系统控制外环控制变量的数值发生变化后系统预期达到的目标参考值，可由从系统给定/用户设定中获得，θ[n]为第[n]个系统采样周期内系统控制外环控制变量的采样值。
需要注意的是，本实施例中，系统控制外环控制变量θ为永磁牵引电机的机械角速度ω m。具体地，本实施例中，外环反馈值与奖励函数中热权重系数的函数表达式为：
Figure PCTCN2021117167-appb-000059
需要注意的是，本实施例中，
Figure PCTCN2021117167-appb-000060
ω start＝352rad/s，ω ref＝530rad/s。
步骤33：构建基于热权重系数动态调整的目标奖励函数，可表示为：
R′(S(l))＝∑λ mG m(S(l))+λ el_dyG el(S(l))。
具体地，本实施例中，基于热权重系数动态调整的奖励函数为：
Figure PCTCN2021117167-appb-000061
步骤S4：在总数为L的牵引变流器各相桥臂电平状态所有可能组合中，选取使得所述目标奖励函数值最大化的系统电平状态组合作为系统控制指令输出，实现相对应功率模块热量的智能调控。具体实现可包括下述步骤。
步骤41：以使基于热权重系数动态调整的奖励函数值最大为优化目标，建立一步优化的计算函数，可表示为：
R′[H]＝max{R′(S(L))}
式中，R′[H]表示当第H组牵引变流器各相桥臂电平状态组合S(H)作用下奖励函数最大化所取到的数值，S(l)∈S＝[S 1,S 2,…,S k,…,S K]；max{·}表示在总数为L的牵引变流器各相桥臂电平状态所有可能组合S＝[S 1,S 2,…,S k,…,S K]中选取使得奖励函数最大化时对应的奖励函数的值。
步骤42：将第H组牵引变流器各相桥臂电平状态组合S(H)作为[n]个系统采样周期内的系统控制指令输出，控制各功率器件的导断状态，从而实现功率模块热量的智能调控。
在第n个系统采样周期，通过系统传感器采样获得i 1[n]、i 2[n]、i 3[n]、θ e[n]和ω m[n]；进一步，通过系统外环控制策略可获得i d_ref[n]和i q_ref[n]；另外，通过系统参考指令/用户设定可获得ω ref和ω start。
具体地，本实施例中，在系统运行在某一稳定速度下，未采用智能延寿控制策略(对应热权重系数λ el＝0)和采用智能延寿控制策略(对应热权重系数λ el≠0)的U相四个功率模块的温度变化，如图4所示。由此可见，在采用智能延寿控制策略的牵引变流器系统的热场分布更为均匀，桥臂各功率模块的温度趋于一致。相比未采用智能延寿控制策略的牵引变流器系统，本专利的控制方法可以是得各功率模块的温度及其波动相近，按照相关研究结果可 知，此情况下各功率模块的寿命消耗也将趋于相近，牵引变流器整体服役寿命将有效避免“木桶短板”效应，从而实现其整体服役寿命的延寿。
具体地，本实施例中，在系统运行加速过程中，随着系统速度逐渐接近目标速度，所述的热权重系数将不断提升；进而，所述基于热权重系数的奖励函数中，热分布控制量对应的归一化控制目标函数的值也随之增大；致使，在一步优化中选出第H组牵引变流器各相桥臂电平状态组合S(H)作用下，系统热分布的情况将得到逐步改善，其热场分布情况逐渐趋于均分，各功率模块的温度及其波动也逐渐趋于一致；最终，实现在列车动态变化中牵引变流器系统热场分布、各功率模块温度的智能调控，进而达到延长牵引变流器系统整体服役寿命的目标。其中，在加速过程中V相4个功率模块的温度情况，如图5所示。
实施例2
与上述方法实施例相对应地，本实施例提供一种热场均匀分布的牵引变流器智能延寿控制系统，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，处理器执行计算机程序时实现上述方法的步骤。
综上，本发明上述两实施例所分别公开的方法及系统，通过智能调控牵引变流器中各功率器件模块产生的热量，实现牵引变流器热场的均匀分布。该方法易于实施，无需额外硬件设备，对延长牵引变流器整体服役寿命，提高列车运行可靠性水平，降低设备维护成本等具有重要意义。
以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (10)

  1. 一种牵引变流器热场控制方法，其特征在于，包括：
    步骤S1：根据各功率模块产生功耗与所属桥臂电平状态和与功耗间接关联的非热场控制量的关系，构建各功率模块对应系统热场控制量的功耗预测模型；
    步骤S2：根据对应系统热场控制量的功耗预测模型构建基于功耗方差的热场分布控制目标函数，然后将基于功耗方差的热场分布控制目标函数与各所述非热场控制量的控制目标函数相结合以构建基于性能的归一化控制目标函数；
    步骤S3：根据所述归一化控制目标函数建立基于性能归一化的初始奖励函数，并将所述初始奖励函数优化为根据外环反馈值动态调整热权重系数的目标奖励函数；
    步骤S4：在总数为L的牵引变流器各相桥臂电平状态所有可能组合中，选取使得所述目标奖励函数值最大化的系统电平状态组合作为系统控制指令输出，实现相对应功率模块热量的智能调控。
  2. 根据权利要求1所述的牵引变流器热场控制方法，其特征在于，所述步骤S2所述的热场分布控制目标函数用于在一段时间内，使得牵引变流器各功率模块总功耗趋于相同，进而产生趋于相近的热量，并使得牵引变流器系统中各功率模块所产生的热量将形成热量分布均匀的一种热系统；所述热场是指牵引变流器运行中各功率模块因自身功耗而产生的热量所形成的热系统。
  3. 根据权利要求1或2所述的牵引变流器热场控制方法，其特征在于，所述非热场控制量为电流。
  4. 根据权利要求3所述的牵引变流器热场控制方法，其特征在于，所述步骤S1具体包括以下步骤：
    S11：建立系统非热场控制量与所属桥臂电平状态的关系，构建系统非热场控制量的预测模型，表示为：
    x m[n+1]＝A m[n]x m[n]+B m(S[n])u[n]+C[n]
    式中，x m[n+1]为预测第[n+1]个系统采样周期内第m类系统非热场控制量的状态向量，x m[n]为第[n]个系统采样周期第m类系统非热场控制量的状态向量，
    Figure PCTCN2021117167-appb-100001
    为与第m类系统非热场控制量有关的状态空间方程中的第p个状态变量，
    Figure PCTCN2021117167-appb-100002
    数值为第m类非热场控制量在第[n]个系统采样周期内的采样值，P为与第m类系统非热场控制量有关的状态空间方程中的状态变量总数，m＝1,2,…,M，M为系统非热场控制量的总数；A m[n]为与第m类系统非热场控制量有关的状态空间方程中与状态变量有 关的时变参数矩阵，B m(S[n])为与第m类系统非热场控制量有关的状态空间方程中的输入变量的时变参数矩阵，其表达式与S[n]有关，S[n]为第[n]个系统采样周期内牵引变流器各相桥臂的电平状态所有可能组合的集合，S＝[S 1,S 2,…,S k,…,S K]，S k表示牵引变流器第k相桥臂的电平状态，其中桥臂的电平状态由若干离散数值组成，具体数量及数值由牵引变流器的拓扑结构所决定，K为牵引变流器的桥臂总数；u[n]为输入向量，u＝[u 1,u 2,…,u q,…,u Q] T，u q为第q个输入变量，Q为输入变量总数；C m为与第m类系统非热场控制量有关的状态空间方程中其它项的时变参数矩阵；
    S12：建立功率模块产生功耗与所属桥臂电平状态的关系，构建基于功耗的系统热场控制量的预测模型，其表达式为：
    Figure PCTCN2021117167-appb-100003
    式中，
    Figure PCTCN2021117167-appb-100004
    为牵引变流器第k相桥臂第j个功率模块在第[n+1]个系统采样周期内的总体功耗预测值，k＝1,2,…,K，j＝1,2,…,J，J为牵引变流器中桥臂功率模块的总数；S k[n]表示在第[n]个系统采样周期内牵引变流器第k相桥臂的电平状态；i k[n+1]为在第[n+1]个系统采样周期内流经牵引变流器第k相桥臂的电流预测值；
    Figure PCTCN2021117167-appb-100005
    和
    Figure PCTCN2021117167-appb-100006
    分别为牵引变流器第k相桥臂第j个功率模块的导通损耗和切换损耗，数值与第[n+1]个系统采样周期内流经牵引变流器第k相桥臂的电流值i k[n+1]和第[n]个系统采样周期内功率模块本身结温T kj[n]有关；
    Figure PCTCN2021117167-appb-100007
    和
    Figure PCTCN2021117167-appb-100008
    分别为判断第j个功率模块是否产生导通损耗和产生切换损耗的函数；其中，i k[n+1]根据S11得到的系统电流控制量的预测值。
  5. 根据权利要求3所述的牵引变流器热场控制方法，其特征在于，所述步骤S2具体包括：
    S21:建立系统非热场控制量的控制目标函数，表示为：
    Figure PCTCN2021117167-appb-100009
    式中，g m[n]为第[n]个系统采样周期内第m类系统非热场控制量的控制目标函数的值，f m为第m类系统非热场控制量的控制目标函数；
    Figure PCTCN2021117167-appb-100010
    为第[n]个系统采样周期内第m类系 统非热场控制量的系统参考给定值/用户设定值组成的向量，
    Figure PCTCN2021117167-appb-100011
    为第m类系统非热场控制量中第p个状态变量的参考值；
    S22：建立基于功耗方差的热场分布控制目标函数，表示为：
    Figure PCTCN2021117167-appb-100012
    式中，g el[n]为第[n]个系统采样周期内系统热场控制量的控制目标函数的值，Var(·)为基于功耗方差的热场分布控制目标函数；
    Figure PCTCN2021117167-appb-100013
    为第[n]个系统采样周期内牵引变流器第k相桥臂第j个功率模块在[n]个系统采样周期前的一个开关周期t sw内的总功耗，t s为系统采样周期；
    S23：构建基于性能的归一化控制目标函数，其统一的表达式为：
    Figure PCTCN2021117167-appb-100014
    式中，G type(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下系统某一控制量type基于性能的归一化控制目标函数的值，其中，G type(S(l))∈(0,1]，l＝1,2，…，L，L为牵引变流器各相桥臂电平状态所有可能组合的总数，系统某一控制量type为系统非热场控制量和系统热场控制量的统称，即type∈{m,el}；
    Figure PCTCN2021117167-appb-100015
    为第l组牵引变流器各相桥臂电平状态组合作用下系统某一控制量type第num_type个状态变量的基于性能的控制目标函数值之和；NUM_type为系统某一控制量type所含状态变量的总数；min{·}表示在总数为L的牵引变流器各相桥臂电平状态所有可能组合S＝[S 1,S 2,…,S k,…,S K]中选取使得控制目标函数最小时对应的控制目标函数的值；
    其中，当type为系统非热场控制量或热场控制量时，根据步骤S21或S22计算得到g m[n]或g el[n]，作为第num_type个状态变量的基于性能的控制目标函数值
    Figure PCTCN2021117167-appb-100016
  6. 根据权利要求5所述的牵引变流器热场控制方法，其特征在于，若所述的归一化表达 式G type(S(l))为1时，表示在第l组牵引变流器各相桥臂电平状态组合作用下系统某一控制量type的控制目标函数能达到最小值，即此时该控制量的性能最优；若所述的归一化表达式G type(S(l))小于1时，G type(S(l))的数值表示在第l组牵引变流器各相桥臂电平状态组合作用下系统某一控制量type的性能为其性能最优时的占比，用百分比表示。
  7. 根据权利要求3所述的牵引变流器热场控制方法，其特征在于，所述步骤S3具体包括以下步骤：
    S31：建立基于性能归一化的初始奖励函数，表示为：
    Figure PCTCN2021117167-appb-100017
    式中，R(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下基于性能归一化的初始奖励函数的值，λ m为第m类系统非热场控制量的归一化控制目标函数G m的权重系数，且λ m取值范围结合实际应用确定，G m(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下第m类系统非热场控制量的归一化控制目标函数的值；λ el为热场控制量的归一化控制目标函数G el的权重系数，且λ el取值范围需结合实际应用确定，G el(S(l))为第l组牵引变流器各相桥臂电平状态组合S(l)作用下热场控制量的基于功耗方差的热场分布归一化控制目标函数的值；
    S32：建立外环反馈值与奖励函数中热权重系数的关系，表示为：
    Figure PCTCN2021117167-appb-100018
    式中，λ el_dy为基于外环反馈值的动态调整热权重系数的数值，f el(·)为描述外环反馈值与奖励函数中热权重系数关系的函数；
    Figure PCTCN2021117167-appb-100019
    为系统控制外环控制变量的数值发生变化前系统/用户所设定的热权重系数的数值，
    Figure PCTCN2021117167-appb-100020
    为系统控制外环控制变量的数值发生变化并到达稳定后系统/用户所设定的热权重系数的数值；θ start为系统控制外环控制变量的数值发生变化前的起始值，θ ref为系统控制外环控制变量的数值发生变化后系统预期达到的目标参考值，θ[n]为第[n]个系统采样周期内系统控制外环控制变量的采样值；
    S33：构建基于热权重系数动态调整的目标奖励函数R′(S(l))，表示为：
    R′(S(l))＝Σλ mG m(S(l))+λ el_dyG el(S(l))。
  8. 根据权利要求7所述的牵引变流器热场控制方法，其特征在于，所述控制外环反馈值包括系统闭环控制中外环控制变量的传感器采样值和系统参考给定/用户的设定值。
  9. 根据权利要求3所述的牵引变流器热场控制方法，其特征在于，所述步骤S4具体包括以下步骤：
    S41：以使基于热权重系数动态调整的奖励函数值最大为优化目标，建立一步优化的计算函数，表示为：
    R′[H]＝max{R′(S(L))}
    式中，R′[H]表示当第H组牵引变流器各相桥臂电平状态组合S(H)作用下奖励函数最大化所取到的数值，S(l)∈S＝[S 1,S 2,…,S k,…,S K]；max{·}表示在总数为L的牵引变流器各相桥臂电平状态所有可能组合S＝[S 1,S 2,…,S k,…,S K]中选取使得奖励函数最大化时对应的奖励函数的值；
    S42：将第H组牵引变流器各相桥臂电平状态组合S(H)作为[n]个系统采样周期内的系统控制指令输出，控制各功率器件的导断状态，从而实现功率模块热量的智能调控。
  10. 一种牵引变流器热场控制系统，包括存储器、处理器以及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述处理器执行所述计算机程序时实现上述权利要求1至9任一所述方法的步骤。
PCT/CN2021/117167 2021-02-01 2021-09-08 牵引变流器热场控制方法及系统 WO2022160723A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110137282.X 2021-02-01
CN202110137282.XA CN112803792B (zh) 2021-02-01 2021-02-01 牵引变流器热场控制方法及系统

Publications (1)

Publication Number Publication Date
WO2022160723A1 true WO2022160723A1 (zh) 2022-08-04

Family

ID=75813368

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/117167 WO2022160723A1 (zh) 2021-02-01 2021-09-08 牵引变流器热场控制方法及系统

Country Status (2)

Country Link
CN (1) CN112803792B (zh)
WO (1) WO2022160723A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115065223A (zh) * 2022-08-18 2022-09-16 杭州飞仕得科技有限公司 一种有源钳位三电平电路的优化控制方法和装置
CN116780929A (zh) * 2023-05-04 2023-09-19 东莞市海柯电子有限公司 基于SiWBG级联H桥变换器的变权重预测控制方法

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112803792B (zh) * 2021-02-01 2022-04-08 中南大学 牵引变流器热场控制方法及系统
CN113381635A (zh) * 2021-06-29 2021-09-10 中南大学 牵引变流器热性能动态优化控制方法及系统

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120221288A1 (en) * 2011-02-28 2012-08-30 General Electric Company, A New York Corporation System and Methods for Improving Power Handling of an Electronic Device
WO2017062913A1 (en) * 2015-10-08 2017-04-13 Johnson Controls Technology Company Photovoltaic energy system with preemptive ramp rate control
CN108880368A (zh) * 2018-06-08 2018-11-23 湖南机电职业技术学院 基于改进型黄金分割法的永磁同步电机效率优化控制方法
CN109742961A (zh) * 2018-11-30 2019-05-10 沈阳工业大学 一种模块化多电平换流器的热平衡控制方法
CN110289781A (zh) * 2019-07-02 2019-09-27 中南大学 一种中点钳位型三电平逆变器温度平衡控制方法及系统
CN112803792A (zh) * 2021-02-01 2021-05-14 中南大学 牵引变流器热场控制方法及系统

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8873644B1 (en) * 2013-07-30 2014-10-28 Texas Instruments Deutschland Gmbh Isolated modulator circuit with synchronized pulse generator for self-monitoring reset with capacitively-coupled isolation barrier
CN108196118A (zh) * 2018-02-14 2018-06-22 武汉大学 一种隔离开关电机输出功率检测装置
CN110719017A (zh) * 2019-09-12 2020-01-21 中车唐山机车车辆有限公司 牵引变流器开关器件限制结温与降低损耗的装置及方法
CN110729911B (zh) * 2019-10-23 2021-06-11 湖南大学 适用于多电平变换器的热应力平衡预测控制方法及系统
CN111371301B (zh) * 2020-03-13 2021-11-02 中南大学 一种两电平牵引逆变器igbt结温控制方法及系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120221288A1 (en) * 2011-02-28 2012-08-30 General Electric Company, A New York Corporation System and Methods for Improving Power Handling of an Electronic Device
WO2017062913A1 (en) * 2015-10-08 2017-04-13 Johnson Controls Technology Company Photovoltaic energy system with preemptive ramp rate control
CN108880368A (zh) * 2018-06-08 2018-11-23 湖南机电职业技术学院 基于改进型黄金分割法的永磁同步电机效率优化控制方法
CN109742961A (zh) * 2018-11-30 2019-05-10 沈阳工业大学 一种模块化多电平换流器的热平衡控制方法
CN110289781A (zh) * 2019-07-02 2019-09-27 中南大学 一种中点钳位型三电平逆变器温度平衡控制方法及系统
CN112803792A (zh) * 2021-02-01 2021-05-14 中南大学 牵引变流器热场控制方法及系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115065223A (zh) * 2022-08-18 2022-09-16 杭州飞仕得科技有限公司 一种有源钳位三电平电路的优化控制方法和装置
CN115065223B (zh) * 2022-08-18 2022-11-29 杭州飞仕得科技有限公司 一种有源钳位三电平电路的优化控制方法和装置
CN116780929A (zh) * 2023-05-04 2023-09-19 东莞市海柯电子有限公司 基于SiWBG级联H桥变换器的变权重预测控制方法
CN116780929B (zh) * 2023-05-04 2024-01-30 东莞市海柯电子有限公司 基于SiWBG级联H桥变换器的变权重预测控制方法

Also Published As

Publication number Publication date
CN112803792A (zh) 2021-05-14
CN112803792B (zh) 2022-04-08

Similar Documents

Publication Publication Date Title
WO2022160723A1 (zh) 牵引变流器热场控制方法及系统
Yin et al. An improved DC-link voltage fast control scheme for a PWM rectifier-inverter system
CN108900119B (zh) 基于死区效应的永磁同步电机模型预测控制方法
Caseiro et al. Cooperative and dynamically weighted model predictive control of a 3-level uninterruptible power supply with improved performance and dynamic response
WO2023035707A1 (zh) 能量变换系统及其过温运行控制方法和控制设备
Zhang et al. Deadbeat control for single-inductor multiple-output DC–DC converter with effectively reduced cross regulation
WO2024125538A1 (zh) 氢燃料电池系统的反馈控制方法、装置及设备
Yin et al. A hybrid modular multilevel converter comprising SiC MOSFET and Si IGBT with its specialized modulation and voltage balancing scheme
CN105720879A (zh) 一种基于直接特征控制的新型双馈感应电机控制方法
CN103715863A (zh) 一种变换器及其功率器件的均热控制方法
Li et al. The impact of VSFPWM on DQ current control and a compensation method
Wang et al. Voltage equalization control of three-phase cascaded H-bridge multilevel rectifiers
WO2023272939A1 (zh) 牵引变流器热性能动态优化控制方法及系统
KR20230054895A (ko) 반직접 구동 직류 풍력 터빈 발전기 유닛, 및 그 제어 방법 및 디바이스
CN106941258B (zh) 一种应用于换流器的功率因数控制方法及装置
Wang et al. Research on optimization method for passive control strategy in CLLC-SMES system based on BP neural network
CN113437895A (zh) 一种矩阵矢量角比例谐振控制方法
Ganesan et al. Coordinated control of rotor side converter and grid side converter on 15 kW DFIG using cascaded feedback linearization controllers
CN114531087B (zh) 一种基于电流源逆变器的高速永磁同步电机优化控制方法
CN110071528B (zh) 一种逆变器并网型分布式电源故障模型构建方法
CN114336660A (zh) 一种基于功角的upqc直接电流预测控制方法
Jin et al. Two-vector predictive current control strategy based on maximum torque per ampere control for PMSMs
CN207339662U (zh) 一种用于高速公路供配电源端电源柜
CN107425546B (zh) 精确调整逆变器母线电压从而提高并网效率的方法
CN109428341B (zh) 一种直流电压协调控制方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21922317

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21922317

Country of ref document: EP

Kind code of ref document: A1