Charging Method, Device and Charger Based on Real-Time Feedback Control

Abstract

The present invention provides a charging method, device and charger based on real-time feedback control, including establishing a communication connection with a power supply device and a device to be charged based on a protocol, and collecting target electrical state quantities of the device to be charged; extracting state data of the device to be charged and state data of the power supply device, constructing a charging prediction control model, and constructing a feedback adjustment model according to the charging prediction control model; iteratively optimizing the charging prediction control model and the feedback adjustment model to generate a charging strategy model; based on the charging strategy model, obtaining a charging strategy, and charging the device to be charged to the target electrical state quantity based on the charging strategy. The present invention achieves the effect of balanced adjustment of the power output quantity between the power supply device and the power supply device, achieves the effect of smooth growth of the power change curve of the device to be charged, and the smoothly increasing power change enables no excessive voltage fluctuation inside the battery, so that on the premise of ensuring the charging efficiency, the minimum loss of the battery is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent charging, and in particular to a charging method, device and charger based on real-time feedback control. Background Art

[0002] The charging process has the greatest impact on the battery life, while the discharging process has less impact. During the charging process, chemical reactions inside the battery may cause voltage fluctuations. That is to say, the vast majority of storage batteries are not damaged by use, but are "charged to death". Thus, it can be seen that a good charger has a very important impact on the service life of the battery.

[0003] At present, there are many electric vehicle chargers on the market, and their performances vary with different manufacturers. Some use linear transformers for step-down and then rectification, some use half-bridge power conversion circuits, and some use single-ended flyback power conversion circuits, etc. However, basically the circuits are relatively simple, the protection functions are poor, and the charging is not carried out according to the charging curve of the battery. Therefore, it may affect the service life of the battery. At the same time, when the battery power changes, it will affect the charging efficiency of the battery. Summary of the Invention

[0004] (I) Technical Problems to be Solved

[0005] In view of the deficiencies of the prior art, the present invention provides a charging method, device and charger based on real-time feedback control to solve the problems raised in the above background art.

[0006] (II) Technical Solutions

[0007] To achieve the above object, the present invention is realized through the following technical solutions: A charging method based on real-time feedback control includes the following steps:

[0008] Establish a communication connection with the power supply device and the device to be charged based on a protocol, and collect the target electrical state quantity of the device to be charged;

[0009] Extract the state data of the device to be charged and the state data of the power supply device, construct a charging prediction control model, and construct a feedback adjustment model according to the charging prediction control model;

[0010] Among them, the state data of the device to be charged includes the pre-charging power data, charging time and charging power change data, and the state data of the power supply device includes the power output of the power supply device;

[0011] Iteratively optimize the charging prediction control model and the feedback adjustment model to generate a charging strategy model;

[0012] Based on the charging strategy model, obtain a charging strategy, and charge the device to be charged to the target electrical state quantity based on the charging strategy.

[0013] Preferably, in this embodiment, the construction of the charging prediction control model includes:

[0014] Calculating the power state curve of the device to be charged during charging and the state quantity S at time t of the device to be charged by using the pre-charging power data, charging time, and charging power change data t ;

[0015] Calculating the power output curve of the power supply device and the supply quantity L at time t of the power supply device by using the power output quantity of the power supply device t ;

[0016] Taking the power state curve of the device to be charged during charging and the state quantity S at time t of the device to be charged t and the power output quantity of the power supply device during power supply and the supply quantity L at time t of the power supply device t as the data training set and inputting it into the decision tree model for training, and outputting the output result of the decision tree model as the charging prediction control model.

[0017] Preferably, in this embodiment, the construction of the feedback regulation model includes:

[0018] Calculating the state quantity S at time t + 1 of the device to be charged by using the supply quantity L at time t of the power supply device t ; t+1 ;

[0019] Calculating the correction function W(L t , S t+1 ) according to the supply quantity L at time t of the power supply device, the state quantity S at time t + 1 of the device to be charged, and the power state change curve of the device to be charged from time t to t + 1 t , S t );

[0020] Calculating the feedback function P(S t+1 , L t , S t ) according to the state quantity S at time t + 1 of the device to be charged, the power output curve of the power supply device from time t to t + 1, and the correction function W(L t+1 , S t ).

[0021] Preferably, in this embodiment, the correction function

[0022] W(L t , S t ) = S t+1 +µL t

[0023] where µ represents the correction coefficient;

[0024] The feedback function P(S t+1 , L t ) = [(S t+1 - S t ) - (S G t+1 - S t )] * γ

[0025] γ represents the feedback coefficient, and S G t+1 represents the state quantity of the device to be charged at the corrected t + 1 moment.

[0026] Preferably, as an embodiment, the iterative optimization of the charging prediction control model and the feedback regulation model includes:

[0027] Supplement the state quantity S t of the device to be charged at time t, the supply quantity L t of the power supply device at time t, the correction function W(L t , S t ) and the feedback function P(S t+1 , L t ) into the corresponding data sets;

[0028] Cluster these data sets respectively, and extract the sample clusters after clustering as training data;

[0029] Train the charging prediction control model and the feedback regulation model respectively based on the training data until the iteration time and the number of times reach the maximum value.

[0030] Preferably, as an embodiment, a charging device based on real-time feedback control for implementing the above-mentioned charging method based on real-time feedback control includes:

[0031] A data interaction module, which establishes a communication connection with the power supply device and the device to be charged based on a protocol, and collects the target electrical state quantity of the device to be charged;

[0032] A model construction module, which extracts the state data of the device to be charged and the state data of the power supply device, constructs a charging prediction control model, and constructs a feedback regulation model according to the charging prediction control model;

[0033] Among them, the state data of the device to be charged includes the pre-charging power data, the charging time, and the charging power change data, and the state data of the power supply device includes the power output quantity of the power supply device;

[0034] A strategy generation module, which iteratively optimizes the charging prediction control model and the feedback regulation model to generate a charging strategy model;

[0035] The charging module obtains a charging strategy based on a charging strategy model and charges the device to be charged to a target power state based on the charging strategy.

[0036] Preferably, in this embodiment, the process of the model construction module constructing a charging prediction control model includes:

[0037] Calculating the power state curve of the device to be charged during charging and the state quantity S at time t of the device to be charged by using the pre-charging power data, charging time, and charging power change data t ;

[0038] Calculating the power output curve of the power supply device and the supply quantity L at time t of the power supply device by using the power output quantity of the power supply device t ;

[0039] Taking the power state curve of the device to be charged during charging and the state quantity S at time t of the device to be charged t and the power output quantity of the power supply device during power supply and the supply quantity L at time t of the power supply device t As a data training set, input it into the decision tree model for training, and output the output result of the decision tree model as a charging prediction control model.

[0040] Preferably, in this embodiment, the process of the model construction module constructing a feedback adjustment model includes:

[0041] Calculating the state quantity S at time t+1 of the device to be charged by using the supply quantity L at time t of the power supply device t ; t+1 ;

[0042] According to the supply quantity L at time t of the power supply device t , the state quantity S at time t+1 of the device to be charged t+1 and the power state change curve of the device to be charged from time t to t+1, calculate the correction function

[0043] W(L t , S t ) = S t+1 +µL t

[0044] µ represents the correction coefficient;

[0045] According to the state quantity S at time t+1 of the device to be charged t+1 , the power output curve of the power supply device from time t to t+1, and the correction function W(L t , S t ), calculate the feedback function

[0046] P(S t+1 , L t) = [(S t+1 - S t ) - (S G t+1 - S t )] * γ

[0047] γ represents the feedback coefficient, and S G t+1 represents the state quantity of the device to be charged at the moment t + 1 after correction.

[0048] As an optimization of this embodiment, the strategy generation module iteratively optimizes the charging prediction control model and the feedback adjustment model, including:

[0049] Supplement the state quantity S of the device to be charged at time t t , the supply quantity L of the power supply device at time t t , the correction function W(L t , S t ) and the feedback function P(S t+1 , L t ) into the corresponding data sets;

[0050] Cluster these data sets respectively, and extract the sample clusters after clustering as training data;

[0051] Based on the training data, train the charging prediction control model and the feedback adjustment model respectively until the iteration time and number of times reach the maximum value.

[0052] As an optimization of this embodiment, a charger includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the above-mentioned charging method based on real-time feedback control.

[0053] (III) Beneficial effects

[0054] The present invention provides a charging method, device and charger based on real-time feedback control, which have the following beneficial effects: by extracting the state data of the device to be charged and the state data of the power supply device, a charging prediction control model is constructed, and a feedback adjustment model is constructed according to the charging prediction control model; the charging prediction control model and the feedback adjustment model are iteratively optimized to generate a charging strategy model; based on the charging strategy model, a charging strategy is obtained, and the device to be charged is charged to the target power state quantity based on the charging strategy. By determining the corresponding change of the power output quantity of the power supply device when the power quantity data of the device to be charged changes smoothly through the relationship between the device to be charged and the power output quantity of the power supply device, the balance adjustment of the power supply device and the power output quantity of the power supply device is achieved, the effect of smooth growth of the power change curve of the device to be charged is achieved, and the stable growth of the power change makes the internal voltage of the battery not have excessive voltage fluctuations, so that the minimum loss of the battery is realized on the premise of ensuring the charging efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 is a flowchart of the charging method based on real-time feedback control of the present invention;

[0056] Figure 2 is a block diagram of the charging device based on real-time feedback control of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] The following describes in detail the embodiments of the present invention. The examples of the embodiments are shown in the drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

[0058] The following disclosure provides many different embodiments or examples for implementing different structures of the present invention. To simplify the disclosure of the present invention, the components and settings of specific examples are described below. Of course, they are only examples and are not intended to limit the present invention. In addition, the present invention may repeat reference numerals and / or reference letters in different examples. Such repetition is for the purpose of simplification and clarity and does not itself indicate the relationship between the various embodiments and / or settings discussed. In addition, the present invention provides examples of various specific processes and materials, but those of ordinary skill in the art can be aware of the application of other processes and / or the use of other materials.

[0059] As Figure 1 shown, an embodiment of the present invention provides a charging method based on real-time feedback control, including the following steps:

[0060] Establish a communication connection with the power supply device and the device to be charged based on the protocol, and collect the target power state quantity of the device to be charged;

[0061] Extract the status data of the device to be charged and the status data of the power supply device, construct a charging prediction control model, and construct a feedback regulation model according to the charging prediction control model;

[0062] Among them, the status data of the device to be charged includes the pre-charging power data, charging time, and charging power change data, and the status data of the power supply device includes the power output of the power supply device;

[0063] Iteratively optimize the charging prediction control model and the feedback regulation model to generate a charging strategy model;

[0064] Based on the charging strategy model, obtain a charging strategy, and charge the device to be charged to the target power state based on the charging strategy.

[0065] Specifically, by collecting the data of the device to be charged and the power supply device, data support is provided for subsequent control. Based on the collected data, a charging prediction control model is constructed. The charging prediction control model is used to show the relationship between the power output of the device to be charged and the power supply device, and further determine the change amount of the power data of the device to be charged over time during charging. At the same time, through the feedback regulation model, the power output of the power supply device is deduced when the power data of the device to be charged changes evenly over time during charging, so as to complete the stable charging of the device to be powered.

[0066] Furthermore, the construction of the charging prediction control model includes:

[0067] Use the pre-charging power data, charging time, and charging power change data to calculate the power state curve of the device to be charged during charging and the state quantity S at time t of the device to be charged t ;

[0068] Use the power output of the power supply device to calculate the power output curve of the power supply device and the supply quantity L at time t of the power supply device t ;

[0069] Take the power state curve of the device to be charged during charging and the state quantity S at time t of the device to be charged t and the power output of the device to be powered during power supply and the supply quantity L at time t of the power supply device t as a data training set and input it into the decision tree model for training, and output the output result of the decision tree model as the charging prediction control model.

[0070] Among them, the charging prediction control model is analyzed according to the historical charging data of the device to be charged. It mainly determines the power state curve of the device to be charged during charging and the state quantity S at time t of the device to be charged according to the pre-charging power data, charging time, and charging power change data t, calculate the power output curve of the power supply device and the supply amount L at time t of the power supply device based on the power output of the power supply device t , after training these data through the decision tree model, the relationship between the power output of the power supply device and the charging data of the device to be charged can be simulated.

[0071] Among them, the charging data of the device to be charged can be charging-related data such as charging time and the change in the power of the device to be charged over time.

[0072] Furthermore, the construction of the feedback adjustment model includes:

[0073] Utilize the supply amount L at time t of the power supply device t Calculate the state quantity S at time t + 1 of the device to be charged t+1 ;

[0074] According to the supply amount L at time t of the power supply device t , the state quantity S at time t + 1 of the device to be charged t+1 and the power state change curve of the device to be charged from time t to t + 1, calculate the correction function W(L t , S t );

[0075] According to the state quantity S at time t + 1 of the device to be charged t+1 , the power output curve of the power supply device from time t to t + 1, and the correction function W(L t , S t ), calculate the feedback function P(S t+1 , L t ).

[0076] Furthermore, the correction function

[0077] W(L t , S t ) = S t+1 +µL t

[0078] µ represents the correction coefficient;

[0079] The feedback function P(S t+1 , L t ) = [(S t+1 -S t )-(S G t+1 -S t )]*γ

[0080] γ represents the feedback coefficient, and S G t+1 represents the state quantity of the device to be charged at time t + 1 after correction.

[0081] Specifically, based on the relationship between the charging data of the device to be charged during charging and the power output of the power supply device, the change in the power output of the power supply device when the charging data of the device to be charged transitions smoothly during charging can be determined.

[0082] Furthermore, the iterative optimization of the charging prediction control model and the feedback adjustment model includes:

[0083] Supplement the state quantity S of the device to be charged at time t t and the supply quantity L of the power supply device at time t t , the correction function W(L t , S t ) and the feedback function P(S t+1 , L t ) into the corresponding data sets;

[0084] Cluster these data sets respectively, and extract the sample clusters after clustering as training data;

[0085] Train the charging prediction control model and the feedback adjustment model respectively based on the training data until the iteration time and number reach the maximum value.

[0086] Among them, through continuous iterative optimization, the charging prediction control model and the feedback adjustment model can be further optimized.

[0087] The charging method based on real-time feedback control of the present invention extracts the state data of the device to be charged and the state data of the power supply device, constructs a charging prediction control model, and constructs a feedback adjustment model according to the charging prediction control model; performs iterative optimization on the charging prediction control model and the feedback adjustment model to generate a charging strategy model; based on the charging strategy model, obtains a charging strategy, and charges the device to be charged to the target power state quantity based on the charging strategy. By the relationship between the device to be charged and the power output of the power supply device, when the power data of the device to be charged changes smoothly, the corresponding change in the power output of the power supply device is determined, achieving the effect of balanced adjustment of the power supply device and the power output of the power supply device, achieving the effect of smooth growth of the power change curve of the device to be charged, and the smoothly increasing power change makes there be no excessive voltage fluctuations inside the battery, so that on the premise of ensuring the charging efficiency, the minimum loss of the battery is achieved.

[0088] As Figure 2 shown, the present invention also provides a charging device based on real-time feedback control for implementing the above-mentioned charging method based on real-time feedback control, including:

[0089] A data interaction module that establishes a communication connection with the power supply device and the device to be charged based on a protocol and collects the target power state quantity of the device to be charged;

[0090] The model construction module extracts the status data of the device to be charged and the status data of the power supply device, constructs a charging prediction and control model, and constructs a feedback adjustment model based on the charging prediction and control model;

[0091] Among them, the status data of the device to be charged includes the power data before charging, the charging time, and the charging power change data, and the status data of the power supply device includes the power output of the power supply device;

[0092] The policy generation module iteratively optimizes the charging prediction and control model and the feedback adjustment model to generate a charging policy model;

[0093] The charging module obtains a charging policy based on the charging policy model, and charges the device to be charged to the target power state based on the charging policy.

[0094] Further, the process of the model construction module constructing the charging prediction and control model includes:

[0095] Using the power data before charging, the charging time, and the charging power change data to calculate the power state curve of the device to be charged during charging and the state quantity S of the device to be charged at time t; t ;

[0096] Using the power output of the power supply device to calculate the power output curve of the power supply device and the supply quantity L of the power supply device at time t; t ;

[0097] Taking the power state curve of the device to be charged during charging and the state quantity S of the device to be charged at time t t and the power output of the power supply device during power supply and the supply quantity L of the power supply device at time t t as a data training set and inputting it into the decision tree model for training, and outputting the output result of the decision tree model as the charging prediction and control model.

[0098] Further, the process of the model construction module constructing the feedback adjustment model includes:

[0099] Using the supply quantity L of the power supply device at time t t to calculate the state quantity S of the device to be charged at time t + 1 t+1 ;

[0100] According to the supply quantity L of the power supply device at time t t , the state quantity S of the device to be charged at time t + 1 t+1 and the power state change curve of the device to be charged from time t to t + 1, calculate the correction function

[0101] W(L t , S t ) = St+1 +µL t

[0102] µ represents a correction coefficient;

[0103] According to the state quantity S at the (t + 1)-th moment of the device to be charged t+1 , the power output curve of the power supply device from time t to t + 1, and the correction function W(L t , S t ), calculate the feedback function

[0104] P(S t+1 , L t ) = [(S t+1 - S t ) - (S G t+1 - S t )] * γ

[0105] γ represents a feedback coefficient, and S G t+1 represents the state quantity of the device to be charged at the (t + 1)-th moment after correction.

[0106] Furthermore, the iterative optimization of the charging prediction control model and the feedback regulation model by the strategy generation module includes:

[0107] Supplement the state quantity S at time t of the device to be charged t , the supply quantity L at time t of the power supply device t , the correction function W(L t , S t ), and the feedback function P(S t+1 , L t ) into the corresponding data sets;

[0108] Cluster these data sets respectively, and extract the sample clusters after clustering as training data;

[0109] Train the charging prediction control model and the feedback regulation model respectively based on the training data until the iteration time and the number of times reach the maximum value.

[0110] The present invention also provides a charger, which includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements the above-mentioned charging method based on real-time feedback control.

[0111] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirits of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A charging method based on real-time feedback control, characterized in that Including the following steps: Establish a communication connection with the power supply device and the device to be charged based on the protocol, and collect the target electrical state quantity of the device to be charged; Extract the state data of the device to be charged and the state data of the power supply device, construct a charging prediction control model, and construct a feedback regulation model according to the charging prediction control model; Among them, the state data of the device to be charged includes the pre-charging power data, charging time, and charging power change data, and the state data of the power supply device includes the power output of the power supply device; Iteratively optimize the charging prediction control model and the feedback regulation model to generate a charging strategy model; Based on the charging strategy model, obtain a charging strategy, and charge the device to be charged to the target electrical state quantity based on the charging strategy; Using the supply quantity L at time t of the power supply device t Calculate the state quantity S at time t + 1 of the device to be charged t+1 ; According to the supply quantity L at time t of the power supply device t , the state quantity S at time t + 1 of the device to be charged t+1 and the power state change curve of the device to be charged from time t to t + 1, calculate the correction function W(L t , S t ); According to the state quantity S at the time t+1 of the device to be charged t+1 , the power output curve of the power supply device from time t to t+1, and the correction function W(L t , S t ), calculate the feedback function P(S t+1 , L t ); The correction function W(L t , S t ) = S t+1 + µL t µ represents a correction coefficient; The feedback function P(S t+1 , L t ) = [(S t+1 - S t ) - (S G t+1 - S t )] * γ γ represents the feedback coefficient, and S G t+1 represents the state quantity of the device to be charged at the corrected time t + 1; The iterative optimization of the charging prediction control model and the feedback regulation model includes: The state quantity S of the device to be charged at time t t , the supply quantity L of the power supply device at time t t , the correction function W(L t , S t ), and the feedback function P(S t+1 , L t ) are supplemented as corresponding data sets; Cluster these data sets respectively, and extract the clustered sample clusters as training data; Train the charging prediction control model and the feedback regulation model respectively based on the training data until the iteration time and number reach the maximum value.

2. The charging method based on real-time feedback control according to claim 1, wherein The construction of the charging prediction control model includes: Calculate the power state curve of the device to be charged and the state quantity S at time t of the device to be charged by using the pre-charging power data, charging time, and charging power change data t ; Calculating the power output curve of a power supply device and the supply quantity L at time t of the power supply device using the power output quantity of the power supply device t ; The power state curve of the device to be charged and the state quantity S at time t of the device to be charged t and the power output of the device to be powered and the supply quantity L at time t of the power supply device t are used as a data training set to be input into the decision tree model for training, and the output result of the decision tree model is output as a charging prediction control model.

3. A charging device based on real-time feedback control, which is used to implement the charging method based on real-time feedback control according to any one of claims 1-2, characterized in that, Including: A data interaction module that establishes a communication connection with the power supply device and the device to be charged based on the protocol, and collects the target electrical state quantity of the device to be charged; A model construction module that extracts the state data of the device to be charged and the state data of the power supply device, constructs a charging prediction control model, and constructs a feedback regulation model according to the charging prediction control model; Among them, the state data of the device to be charged includes the pre-charging power data, charging time, and charging power change data, and the state data of the power supply device includes the power output of the power supply device; A strategy generation module that iteratively optimizes the charging prediction control model and the feedback regulation model to generate a charging strategy model; A charging module that obtains a charging strategy based on the charging strategy model, and charges the device to be charged to the target electrical state quantity based on the charging strategy.

4. The charging device based on real-time feedback control according to claim 3, characterized in that The process by which the model construction module constructs the charging prediction control model includes: Calculate the power state curve of the device to be charged and the state quantity S at time t of the device to be charged by using the power data before charging, the charging time, and the charging power change data t ; Calculating the power output curve of a power supply device and the supply amount L at time t of the power supply device using the power output amount of the power supply device t ; The power state curve of the device to be charged and the state quantity S at time t of the device to be charged t and the power output of the power supply device and the supply quantity L at time t of the power supply device t are used as a data training set to be input into the decision tree model for training, and the output result of the decision tree model is output as a charging prediction control model.

5. The charging device based on real-time feedback control according to claim 4, characterized in that, The process by which the model construction module constructs the feedback regulation model includes: Using the supply quantity L at time t of the power supply device t Calculate the state quantity S at time t + 1 of the device to be charged t+1 ; According to the supply amount L at time t of the power supply device t , the state quantity S at time t + 1 of the device to be charged t+1 and the power state change curve of the device to be charged from time t to t + 1, calculate the correction function W(L t , S t ) = S t+1 + µL t µ represents a correction coefficient; According to the state quantity S at the time t+1 of the device to be charged t+1 , the power output curve of the power supply device from time t to t+1, and the correction function W(L t , S t ), calculate the feedback function P(S t+1 , L t ) = [(S t+1 - S t ) - (S G t+1 - S t )] * γ γ represents the feedback coefficient, and S G t+1 represents the state quantity of the device to be charged at the corrected time t + 1.

6. The charging device based on real-time feedback control according to claim 5, wherein The iterative optimization of the charging prediction control model and the feedback regulation model by the strategy generation module includes: The state quantity S of the device to be charged at time t t , the supply quantity L of the power supply device at time t t , the correction function W(L t , S t ), and the feedback function P(S t+1 , L t ) are supplemented as corresponding data sets; Cluster these data sets respectively, and extract the clustered sample clusters as training data; Train the charging prediction control model and the feedback regulation model respectively based on the training data until the iteration time and number reach the maximum value.

7. A charger, the charger comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the charging method based on real-time feedback control according to any one of claims 1-2 when executing the program.

Images