CN-119966264-B - Improved model-free current prediction control method, device and system based on forgetting factor

CN119966264BCN 119966264 BCN119966264 BCN 119966264BCN-119966264-B

Abstract

The invention discloses an improved model-free current prediction control method, device and system based on forgetting factors. The method does not need to acquire system parameters in advance, only needs to acquire load current and approximately calculate the current increment by historical current data, then outputs the optimal inverter switching state by minimizing a current cost function, and simultaneously introduces forgetting factors to weaken the adverse effect of sampling errors on the system, so that model-free current prediction control can be realized. On the basis of ensuring the realization of parameter-free control, the invention introduces forgetting factors to weaken the adverse effect of sampling errors on the system, improves the current quality while having higher response speed, and improves the control robustness. The invention is suitable for different types of power electronic topological structures, does not need to carry out mathematical modeling on the different types of power electronic topological structures independently, has reference significance for the control of model-free prediction current of the power electronic converter, and has wide application prospect.

Inventors

MA ZHIXUN
ZHANG XU
WANG RUOLIN
CHEN XINBO
LIN GUOBIN

Assignees

同济大学

Dates

Publication Date: 20260508
Application Date: 20250125

Claims (9)

1. The improved model-free current prediction control method based on the forgetting factor is characterized by comprising the following steps of: Acquisition in real time of The state of the inverter at the moment is stored as a history data record, and the state of the inverter comprises the first Three-phase current at moment and the first The switch state of the inverter at the moment Time three-phase current representation 、、 ; The first to be collected Clark conversion is carried out on the three-phase current at the moment to obtain the current of the inverter in an alpha-beta static coordinate system; Definition of the first embodiment The current increment at the moment is the first Time and the first The current difference at the moment, and the two adjacent current increment under the same switch state are approximately equal; calculating a current increment using the historical current data; Introducing forgetting factors to update current increment, and calculating predicted current according to the updated current increment; introducing forgetting factor to update current increment, and updating the updated current increment Expressed in general form as: Wherein, the As a forgetting factor, For last switching state of The current increment in the lower part of the current, The selected switch state; obtaining a current cost function according to the relation between the current predicted value and the current reference value; And calculating a current cost function of all the switch states of the inverter, and selecting a switch state corresponding to the minimized current cost function as output.
2. The improved model-free current prediction control method based on forgetting factor as set forth in claim 1, wherein based on the obtained first Time of day current and defined first Time of day, the first The current increment at the moment, the general mathematical expression for current prediction is obtained as follows: Wherein, the Represents alpha, beta axis current; For the inverter switching state at the kth instant, Is the first Sampling value of current at moment; 、 Respectively calculated according to the historical current data Time of day and the first Time of day current increment.
3. The method for improved model-free current predictive control based on forgetting factor as set forth in claim 2, wherein the current delta is calculated using historical current data, the first Time of day and the first Time of day current increment 、 The general form is expressed as: Wherein, the 、、 Respectively the first Time of day, the first Time of day and the first The current sample at the moment.
4. The method for improved model-free current predictive control based on forgetting factor as set forth in claim 3, wherein two adjacent current increments are approximately equal in the same switching state, the first Time of day current increment Expressed in general form as: Wherein, the For last switching state of The current increment in the lower part of the current, 、 Respectively the last switching state is The current collection value at the lower part and the current collection value at the last moment.
5. The improved model-free current prediction control method based on forgetting factor as set forth in claim 4, wherein the predicted current is calculated from the updated current increment in the general form expressed as: 。
6. The method for improved model-free current predictive control based on a forgetting factor as set forth in claim 3, wherein said current cost function is expressed in a generalized form as: Wherein, the 、 The reference currents of the alpha and beta axes respectively, 、 Respectively select the switch states as Time No The predicted current of the alpha, beta axis at the moment, To select the switch state as A function of the current value at the time, In order to select the minimum element function, As a minimum value of the current cost function, Is the first And outputting the optimal inverter switching state at the moment.
7. An improved model-free current prediction control device based on forgetting factors is characterized by comprising: The acquisition and storage module is used for acquiring the first data in real time The state of the inverter at the moment is stored as a history data record, and the state of the inverter comprises the first Three-phase current at moment and the first The switch state of the inverter at the moment Time three-phase current representation 、、 ; Mathematical model module in alpha-beta static coordinate system for collecting three-phase current at kth time 、、 Clark conversion is carried out to obtain the current of the inverter in an alpha-beta static coordinate system; an increment calculation module for defining the first Time of day and the first The current increment at the moment is the current difference between the front moment and the rear moment, and the two adjacent current increments in the same switch state are approximately equal; The current increment updating module is used for introducing forgetting factors to update current increment and calculating the updated current increment to obtain predicted current; introducing forgetting factor to update current increment, and updating the updated current increment Expressed in general form as: Wherein, the As a forgetting factor, For last switching state of The current increment in the lower part of the current, The selected switch state; the current cost function construction module is used for obtaining a current cost function according to the relation between the current predicted value and the current reference value; And the output module is used for calculating the current cost function of the inverter in all the switch states and selecting the switch state corresponding to the minimized current cost function as output.
8. A computer program product comprising computer programs/instructions which, when executed by a processor, implement the improved model-free current predictive control method based on forgetting factors as claimed in any one of claims 1 to 6.
9. An electronic device comprising a processor and a memory, wherein the processor implements the improved model-free current predictive control method based on forgetting factors of any one of claims 1 to 6 when executing a computer program stored in the memory.

Description

Improved model-free current prediction control method, device and system based on forgetting factor Technical Field The invention belongs to the field of power electronic control, and particularly relates to an improved model-free current prediction control method, device and system based on forgetting factors. Background With the rapid development of power electronics technology, power electronic converters are increasingly used in various power electronic systems, and the control strategy of the power electronic converters directly affects the performance and efficiency of the systems. The conventional control method generally relies on an accurate system parameter model for control design, however, in practical application, system parameters are often difficult to accurately acquire, and the accuracy of model parameters is difficult to ensure due to dynamic changes of system operation conditions, which results in insufficient accuracy and robustness of the conventional control method when coping with system parameter changes and sampling errors. The existing model-free predictive current control (MFPCC) method, although only based on detecting load current and current difference without system parameters, improves the accuracy and robustness of control to a certain extent, the sensitivity to sampling errors still limits the application of the method in a high-frequency power electronic system. Therefore, the current prediction control method which can effectively weaken the influence of sampling errors and adapt to the change of the operation conditions without depending on a system parameter model is further researched, and the method has important significance for improving the control performance and the system stability of the power electronic converter. Disclosure of Invention The invention aims to provide an improved model-free current prediction control method, device and system based on forgetting factors, which are used for solving the problems. In order to achieve the above object, the embodiment of the present invention provides the following technical solutions: in a first aspect, the present invention provides an improved model-free current prediction control method based on a forgetting factor, the method comprising: Collecting the state of an inverter at the kth moment in real time and storing the state as a historical data record, wherein the state of the inverter comprises a three-phase current at the kth moment and an inverter switching state at the kth-1 moment, and the three-phase current at the kth moment is expressed as Clark conversion is carried out on the collected three-phase current at the kth moment to obtain the current of the inverter in the alpha-beta static coordinate system; defining the current increment at the kth time as the current difference between the kth time and the kth-1 time, and enabling two adjacent current increments to be approximately equal in the same switching state; calculating a current increment using the historical current data; Introducing forgetting factors to update current increment, and calculating predicted current according to the updated current increment; obtaining a current cost function according to the relation between the current predicted value and the current reference value; And calculating a current cost function of all the switch states of the inverter, and selecting a switch state corresponding to the minimized current cost function as output. In one embodiment, according to the acquired kth current and the defined kth current increment and kth+1 current increment, a general mathematical expression of current prediction is obtained as follows: Wherein i= [ i αiβ]T ] represents alpha, beta axis current, V k is the switch state of the inverter at the kth moment, and the switch state V epsilon { V 0,V1,…,Vm }; a sampling value of the current at the kth moment; the current increment at the kth time and the (k+1) th time are calculated according to the historical current data respectively. In one embodiment, the historical current data is used to calculate the current delta, the current delta at the kth time and the (k+1) th timeThe general form is expressed as: Wherein, the The current sampling values at the kth time, the (k+1) th time and the (k+2) th time are respectively obtained. In one embodiment, two adjacent current increments are approximately equal in the same switching state, the current increment at the kth timeThe general form is expressed as: Wherein, the For the current increment at the last switching state V k,The current collection value of the last switching state is V k and the current collection value of the last moment. In one embodiment, a forgetting factor is introduced to update the current delta, the updated current deltaExpressed in general form as: wherein, eta is a forgetting factor, For the current increment at the last switching state V j,To introduce a current increment updated by a forgetting factor.