Lithium ion battery fast charging heat regulation and control method based on neural network enhanced model predictive control

Abstract

The invention relates to the technical field of temperature control, in particular to a lithium ion battery fast charge heat regulation and control method based on neural network enhanced model predictive control, which comprises the following steps: acquiring a battery signal and calculating a variable quantity, predicting temperature and judging an error grade and trend, calculating a correction value and comparing the correction value with an upper limit to form a regulation set; generating a correction sequence, writing the correction sequence into a sending frame to obtain a set frame set, and recording actual parameter calculation execution deviation to form a deviation record set. An offset combined value mechanism is introduced into adjustment calculation, the adjustment amplitude adaptability is guaranteed, the thermal shock risk is reduced, control instruction setting is combined with a rated value for amplitude verification, execution safety is enhanced, a feedback closed loop is constructed by recording execution deviation and a temperature value, and synchronous improvement of control precision and system adaptability is achieved. A high-dynamic temperature control adjusting mechanism which is driven by a time sequence, quick in response, flexible in adjustment and safe and closed in loop is integrally formed.

Description

Technical Field

[0001] This invention relates to the field of temperature control technology, and in particular to a method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control. Background Technology

[0002] The field of temperature control technology mainly involves technologies related to regulating heat transfer and temperature changes in physical systems. Core aspects include temperature detection, heat regulation, the design and implementation of thermal management systems, and the formulation and optimization of temperature control strategies. This technology is widely used in various industrial scenarios such as electronic equipment, power systems, automobiles, and battery management. The overall technology typically achieves stable temperature control of the system through methods such as establishing thermal models, configuring sensors, designing heating or cooling structures, and implementing control algorithms. Traditional thermal regulation methods for lithium-ion batteries during fast charging address the risk of thermal runaway caused by the large amount of heat generated during high-rate charging. These methods often employ predictive control by embedding a battery thermal management system and combining it with thermal modeling of battery material parameters. These methods typically use predictive control strategies based on physical heat conduction models or temperature control threshold settings based on empirical rules to perform thermal regulation operations. In contrast, the neural network-enhanced model predictive control method for lithium-ion battery fast charging improves the accuracy of predicting future temperature changes and achieves feedforward optimization of the thermal control signal to achieve the temperature control target during fast charging by introducing neural networks to model the battery's thermal behavior.

[0003] Traditional methods rely on static thermal models or heat conduction models based on physical formulas. In actual fast charging, these models struggle to reflect real-time dynamic changes in complex charging environments and lack the ability to identify temperature fluctuation trends and error levels, leading to lag in thermal regulation responses. In practice, when the battery temperature rise rate suddenly increases, the temperature control system often fails to promptly identify subtle temperature changes and trends, potentially delaying intervention and causing the risk of exceeding temperature thresholds. Furthermore, existing solutions often use empirically set control thresholds, lacking adaptive linkage with actual operating data. Regulation commands are often executed in a fixed pattern, easily resulting in excessive or insufficient adjustment, leading to energy waste or lag. During communication link transmission, the control signal lacks dynamic assessment of the matching relationship between the executed value and the rated value, posing a risk of execution failure or hardware overload due to unreasonable command strength. At the feedback control level, there is often a lack of continuous tracking and deviation recording mechanisms for the adjustment effect, preventing the formation of a systematic self-optimizing control chain. This makes it difficult to continuously optimize control results based on actual conditions, weakening regulation accuracy and system robustness. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a method for fast-charging thermal regulation of lithium-ion batteries using neural network enhancement model predictive control, comprising the following steps: S1: Acquire the temperature, current, and voltage signals of the battery module and complete the numerical conversion, calculate the temperature and current changes, and combine the results with the sampling time in sequence to form a fast charging thermal behavior sequence; S2: Based on the fast charging thermal behavior sequence, the main control chip calculates the predicted temperature and error, determines the temperature error level according to the preset interval, obtains the error development trend according to the temperature change direction, and generates a thermal prediction error hierarchical set in chronological order. S3: Based on the thermal prediction error hierarchical set, calculate the current and cooling correction base values ​​respectively, compare them with the corresponding upper limit of amplitude, generate the adjustment content set in time order, and obtain the combined value of charging offset; S4: Based on the combined value of charging and cooling offset, a current and cooling correction setting sequence is generated in the battery management structure, and after comparing it with the rated output current and heat dissipation power, the processing sequence is written into the vehicle communication transmission frame to form a correction control setting frame set. S5: Based on the modified control setting frame set, record the actual output current, heat dissipation power and battery temperature, calculate the current and cooling execution deviation, and combine them with the temperature according to the period to form an execution deviation change record set.

[0005] As a further aspect of the present invention, the fast charging thermal behavior sequence includes temperature change, current change, sampling timestamp, and timing record item number; the thermal prediction error classification set includes predicted temperature value, prediction error value, temperature error interval level, error development trend label, and classification timing index; the charging-cooling offset combination value includes current correction base value, cooling correction base value, current adjustable amplitude judgment result, cooling power adjustable amplitude judgment result, and offset timing entry; the correction control setting frame set includes current correction setting sequence, cooling correction setting sequence, transmission frame field set, link transmission frame identifier, and frame sequence number timing set; and the execution deviation change record set includes actual output current record, actual heat dissipation power record, battery module temperature record, current execution deviation value, and cooling execution deviation value.

[0006] As a further aspect of the present invention, the specific steps of S1 are as follows: S101: Acquire temperature, current and voltage signals from the lithium-ion battery module, input the signals into the battery management structure, perform numerical conversion according to the sensor sampling frequency, form temperature, current and voltage data arranged in the order of sampling time, and generate a sequence of sampled signal values. S102: Based on the numerical sequence of the sampled signal, calculate the temperature change and current change at adjacent sampling times, construct time difference units by combining time tags, arrange them in time order, and obtain the temperature and current change sequence. S103: Based on the temperature and current change sequence, extract the voltage value, temperature change and current change corresponding to the time node, construct a continuous set including time, voltage and change, and generate a fast charging thermal behavior sequence.

[0007] As a further aspect of the present invention, the specific steps of S2 are as follows: S201: Based on the fast charging thermal behavior sequence, extract sampling point information in the main control chip according to the time index, call the temperature change, voltage value and current change of the corresponding time node, combine with the thermal estimation module of the control chip to generate the predicted temperature value of the time point, and arrange them in time order to generate a predicted temperature value sequence. S202: Based on the predicted temperature value sequence, call the actual temperature value at the corresponding time point, calculate the error value between the predicted temperature value and the actual temperature value, compare the error value with the set thermal safety error range, mark the corresponding range level label according to the judgment result, and obtain the temperature error range level sequence. S203: Call the level labels of the time nodes in the temperature error interval level sequence, and compare the differences with the direction of temperature change within the same time point to determine the error change trend. Then, pair the trend value with time and merge it into the interval level sequence to generate a thermal prediction error level set.

[0008] As a further aspect of the present invention, the specific steps of S3 are as follows: S301: Call the temperature error interval level and current change of the time node in the thermal prediction error classification set, calculate the current correction base value at the time point according to the adjustment parameter corresponding to the error level, and generate the current correction base value sequence. S302: Based on the temperature error range level and temperature change amount of the time nodes in the heat prediction error classification set, and combined with the cooling parameters corresponding to the level, calculate the basic value of cooling correction at the time point, and generate a sequence of basic values ​​of cooling correction amount. S303: Call the current correction base value sequence and the cooling correction base value sequence, combine them with the corresponding current adjustable amplitude limit and cooling power adjustable amplitude limit, determine whether the base value exceeds the set amplitude, and arrange them in time order to form an adjustment set, generating a charging offset combination value.

[0009] As a further aspect of the present invention, the specific steps of S4 are as follows: S401: Based on the combined value of charging and cooling offsets, the current offset and cooling offset parameters of the time period are matched in the battery management structure, the corresponding adjustment fields are extracted in sequence, and the fields are structurally matched with the offset combination values ​​and written into the preset field sequence to generate the current correction setting sequence and the cooling correction setting sequence. S402: Call the current correction setting sequence and cooling correction setting sequence, and determine whether the sequence value exceeds the corresponding rated amplitude limit according to the rated value of the output current of the DC-DC converter circuit and the rated value of the heat dissipation power of the liquid cooling plate. Perform amplitude compression processing on the sequence value that exceeds the limit to generate a set of correction setting sequences after amplitude limiting. S403: Based on the set of correction settings after the amplitude limit, extract the sequence values ​​according to the field definition specifications of the vehicle internal communication link, assemble the processed field contents into a multi-frame communication structure in sequence, and write the number and format according to the frame structure settings to generate a set of correction control settings frames.

[0010] As a further aspect of the present invention, the specific steps of S5 are as follows: S501: According to the modified control setting frame set, the output current of the DC-DC converter circuit, the heat dissipation power of the liquid cooling plate and the temperature value of the battery module are recorded sequentially in the battery management structure, and the signals are sampled periodically and organized into a set of periodic operating state values. S502: Based on the current and cooling power signals in the set of periodic operating state values, calculate the current deviation and cooling deviation, and average the deviation values ​​in each period to obtain the current execution deviation value and cooling execution deviation value. S503: Combining the current execution deviation value, the cooling execution deviation value, and the temperature data in the set of periodic operating status values, the deviation and temperature value are paired in chronological order to generate a set of execution deviation change records.

[0011] As a further embodiment of the present invention, the battery module is a packaged unit composed of multiple lithium-ion cells, and integrates temperature, voltage and current sensors. The fast charging thermal behavior sequence is a multi-dimensional time series set composed of temperature changes, current changes, and sampling time obtained by the battery module during fast charging, arranged in order of sampling time.

[0012] As a further aspect of the present invention, the main control chip is the core processing unit in the battery management structure. It is an embedded computing chip of MCU, DSP or FPGA, responsible for performing thermal behavior prediction, error judgment and control strategy calculation operations. The predicted temperature and error are the difference between the predicted temperature value calculated by the main control chip and the sampled actual temperature value. The temperature error level is a classification based on the comparison between the predicted error value and the thermal safety error range, indicating the severity of the current temperature deviation. The error development trend is the future evolution trend of the error determined based on the direction of temperature change, which is used to assist the control strategy in determining whether the correction value needs to be adjusted. The thermal prediction error classification set is a set arranged in chronological order, including temperature error interval levels and error development trends; The base value of the current correction is a reference correction value generated by a calculation model based on the temperature error range level and the current change. The base value for cooling correction is a reference correction value generated based on the temperature error range level and the amount of temperature change. The upper limit of amplitude is a set of control target values ​​formed by combining the maximum current adjustment range of the battery system or DC-DC converter circuit, the base value of the current correction amount, the base value of the cooling correction amount, and the upper limit of the adjustable amplitude.

[0013] As a further aspect of the present invention, the current correction setting sequence is a set of time series setting values ​​that are generated based on the combined values ​​of the charging and cooling offset and indicate the output current adjustment strategy of the DC-DC converter circuit. The cooling correction setting sequence is a time series set of setting values ​​that indicate the liquid cooling plate cooling power adjustment strategy, generated based on the combined values ​​of the cooling offset. The transmitted frame is a frame format data packet transmitted via the vehicle communication link, including a current and cooling correction setting value field. The correction control setting frame set is a collection of multiple transmission frames arranged in chronological order, used to periodically transmit the currently calculated correction control parameters; The current execution deviation refers to the difference between the actual output current of the DC-DC converter circuit and the set value, which is used to provide feedback on the control effect. The cooling performance deviation refers to the difference between the actual heat dissipation power of the liquid cooling plate and the cooling set value; The execution deviation change record set is a collection of records arranged in chronological order, consisting of periodically collected current execution deviation, cooling execution deviation, and battery module temperature values.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, during fast charging, temperature, current, and voltage signals are collected and a thermal behavior sequence is constructed to enhance the temporal correlation of thermal changes. In the prediction stage, the sensitivity of temperature control response is improved by judging the error level and development trend. An offset combination value mechanism is introduced in the adjustment calculation to ensure the adaptability of the adjustment range and reduce the risk of thermal shock. The control command setting is combined with the rated value for amplitude verification to enhance execution safety. By recording the execution deviation and temperature value, a feedback closed loop is constructed to achieve simultaneous improvement in control accuracy and system adaptability. Overall, a highly dynamic temperature control adjustment mechanism with time sequence drive, fast response, flexible adjustment, and safe closed loop is formed. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the steps of the present invention; Figure 2 This is a detailed schematic diagram of S1 of the present invention; Figure 3 This is a detailed schematic diagram of S2 of the present invention; Figure 4 This is a detailed schematic diagram of S3 of the present invention; Figure 5 This is a detailed schematic diagram of S4 of the present invention; Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0017] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0018] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0019] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0020] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0021] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0022] Please see Figure 1 This invention provides a method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control, comprising the following steps: S1: Acquire temperature, current and voltage signals from the lithium-ion battery module and convert them into corresponding values ​​in the battery management structure. Calculate the temperature and current changes, combine the values ​​with the sampling time to form a set of records arranged in chronological order, and obtain the fast charging thermal behavior sequence. A lithium-ion battery module is a packaged unit composed of multiple lithium-ion cells, which integrates temperature, voltage and current sensors to collect data related to the operating status. It is the basic structure that constitutes the object of the fast charging thermal behavior analysis in this invention. Temperature, current and voltage signals are raw physical quantity data acquired by sensors or sampling chips in the battery module to reflect the current operating status of the battery, and are transmitted to the battery management structure via CAN bus or sampling module. The battery management architecture is a system structure consisting of a main control chip, a data processing unit, a current control module, a cooling management module, and a communication interface, used to perform data acquisition, calculation, control, and communication functions; The fast charging thermal behavior sequence is a multi-dimensional time series set consisting of temperature changes, current changes, and sampling time obtained by the battery module during fast charging. It is arranged in order of sampling time and is used for thermal characteristic modeling and predictive analysis. S2: Based on the fast charging thermal behavior sequence, the predicted temperature value of the sampling point is obtained in the main control chip, and then the prediction error value is calculated. The predicted error value is compared with the preset thermal safety error range to obtain the temperature error range level. The error development trend is obtained according to the direction of temperature change, and the thermal prediction error hierarchical set is formed by arranging them in time order. The main control chip is the core processing unit in the battery management structure. It is an embedded computing chip such as MCU, DSP or FPGA, responsible for performing thermal behavior prediction, error judgment and control strategy calculation operations. The prediction error value is the difference between the predicted temperature value calculated by the main control chip and the sampled actual temperature value. It is used to measure the prediction accuracy and guide the correction strategy. The thermal safety error range is a temperature prediction error range preset according to battery manufacturing standards or fast charging safety specifications, including a range of ±2℃, used to determine the safety level of the current prediction value; Temperature error range levels are classifications based on the comparison between predicted error values ​​and thermal safety error ranges, indicating the severity of the current temperature deviation, including normal, warning, and danger ranges. Error development trend is the future evolution trend of error determined based on the direction of temperature change (increasing or decreasing), which is used to assist the control strategy in determining whether the correction value needs to be adjusted. The thermal prediction error hierarchy set is a set arranged in chronological order, including temperature error range levels and error development trends, used to determine the evolution of thermal state during fast charging. S3: Call the thermal prediction error classification set to calculate the base value of the current correction amount in the main control chip according to the temperature error range level and the current change amount. Calculate the base value of the cooling correction amount according to the temperature error range level and the temperature change amount. Then, compare the base value with the preset upper limit of the adjustable current amplitude and the upper limit of the adjustable cooling power amplitude to form a set of adjustment contents arranged in time order, and obtain the combined value of the charging offset amount. The base value of the current correction is a reference correction value generated by a calculation model based on the temperature error range level and the amount of current change. It is used to indicate the direction and magnitude of the current amplitude that needs to be adjusted. The base value for cooling correction is a reference correction value generated based on the temperature error range level and the amount of temperature change. It is used to indicate the extent to which the cooling power of the liquid cooling plate needs to be increased or decreased. The upper limit of the adjustable current amplitude is one of the limiting conditions for the correction amount, which is the maximum current adjustment range of the battery system or DC-DC converter circuit. The upper limit of the adjustable cooling power is the maximum power range that the liquid cooling plate can adjust, which is used to limit the excessive adjustment range of the base value of the cooling correction amount. The combined value of the charging and cooling offset is a set of control target values ​​formed by combining the base value of the current correction and the base value of the cooling correction with the upper limit of the adjustable amplitude. It is arranged in chronological order and used to generate the control setting sequence. S4: Based on the combined values ​​of charging and cooling offsets, generate a current correction setting sequence and a cooling correction setting sequence in the battery management structure, and determine the amplitude of the output current rating of the DC-DC converter circuit and the heat dissipation power rating of the liquid cooling plate. Write the processed sequence into the transmission frame of the vehicle internal communication link according to the specified fields to obtain the correction control setting frame set. The current correction setting sequence is a set of time-series setting values ​​generated based on the combined values ​​of the charging and cooling offset, indicating the output current adjustment strategy of the DC-DC converter circuit. The cooling correction setting sequence is a time-series set of setting values ​​that indicate the liquid-cooled plate cooling power adjustment strategy, generated based on the combined values ​​of the cooling offset. A DC-DC converter circuit is a power regulation circuit used to convert an input DC voltage into a voltage and current suitable for battery charging. It includes a DC / DC converter and is one of the execution components in the battery charging process. Liquid cooling plates are cooling system components installed on the surface of battery modules for heat exchange. The key is to remove battery heat through liquid flow. The vehicle internal communication link refers to the data bus used for inter-module communication in the vehicle system, including CAN (Controller Area Network) or LIN (Local Interconnect Network), which is used to transmit setting frames and monitoring data; The frame format data packet, which is transmitted via the vehicle communication link, includes current and cooling correction setpoint fields to control the behavior adjustment of the actuators; The correction control setting frame set is a collection of multiple transmission frames arranged in chronological order, used to periodically transmit the currently calculated correction control parameters; S5: Based on the correction control setting frame set, the actual output current of the DC-DC converter circuit, the actual heat dissipation power of the liquid cooling plate and the temperature value of the battery module are recorded in the battery management structure. The current execution deviation and cooling execution deviation are calculated. The deviation values ​​and temperature values ​​are arranged in a periodic manner to form a set of change records, and the execution deviation change record set is obtained. Current execution deviation refers to the difference between the actual output current of the DC-DC converter circuit and the set value, which is used to provide feedback on the control effect. Cooling execution deviation refers to the difference between the actual heat dissipation power of the liquid cooling plate and the cooling set value, which is used to evaluate the accuracy of cooling control. The execution deviation change record set is a collection of records arranged in chronological order, consisting of periodically collected current execution deviation, cooling execution deviation, and battery module temperature values, used for control optimization or safety assessment.

[0023] The fast charging thermal behavior sequence includes temperature change, current change, sampling timestamp, and timing record item number; the thermal prediction error classification set includes predicted temperature value, prediction error value, temperature error interval level, error development trend label, and classification timing index; the charging-cooling offset combination value includes current correction base value, cooling correction base value, current adjustable amplitude judgment result, cooling power adjustable amplitude judgment result, and offset timing entry; the correction control setting frame set includes current correction setting sequence, cooling correction setting sequence, transmission frame field set, link transmission frame identifier, and frame sequence number timing set; the execution deviation change record set includes actual output current record, actual heat dissipation power record, battery module temperature record, current execution deviation value, and cooling execution deviation value.

[0024] Please see Figure 2 The specific steps of S1 are as follows: S101: Acquire temperature, current and voltage signals from the lithium-ion battery module, input the signals into the battery management structure, perform numerical conversion according to the sensor sampling frequency, form temperature, current and voltage data arranged in the order of sampling time, and generate a sequence of sampled signal values. Temperature, current, and voltage signals from the lithium-ion battery module are collected by sensors. These signals are processed by the battery management system (BMS). The sensors periodically collect battery operating status data at a sampling frequency of 100 times per second. The collected analog signals are converted into digital signals by an analog-to-digital converter (ADC). These digital signals are transmitted to the BMS, which arranges the sampled signals in time-stamp order to form time-series data of temperature, current, and voltage. During the sampling process, temperature, current, and voltage are recorded at regular intervals according to the sampling frequency. For example, if the sampling interval is 0.01 seconds, data such as 30°C, 5A, and 4.2V will be recorded and a data sequence will be generated according to the sampling order.

[0025] S102: Based on the numerical sequence of the sampled signal, calculate the temperature change and current change at adjacent sampling times, construct time difference units by combining time labels, arrange them in time order, and obtain the temperature and current change sequence; First, by comparing the temperature and current values ​​at adjacent time points, the changes in temperature and current are determined. For example, assuming the temperature is 30℃ and the current is 5A at time 0 seconds, and the temperature is 31℃ and the current is 5.2A at time 0.01 seconds, then the temperature change is 1℃ and the current change is 0.2A. Through this process, the changes in temperature and current at each sampling time can be obtained. These changes, along with time stamps, form time difference units, which are then arranged chronologically to form a time series of temperature and current changes.

[0026] S103: Based on the temperature and current change sequence, extract the voltage value, temperature change and current change corresponding to the time node, construct a continuous set including time, voltage and change, and generate a fast charging thermal behavior sequence; Based on the aforementioned temperature and current change sequence, the voltage value and its change at each time point can be obtained from the collected data. The battery's voltage, temperature, and current change over time. During fast charging, the battery's temperature and current typically increase gradually. By recording these changes and combining them with time stamps, a set containing time, voltage, temperature, and current changes is formed. For example, at a certain moment, the battery voltage is 4.2V, the temperature is 30℃, and the current is 5A. After a certain period, the battery temperature and current increase, the battery voltage remains at 4.2V, the temperature rises to 31℃, and the current increases to 5.2A. At this point, the set of time, voltage, and change (e.g., 0.01 seconds, 4.2V, 1℃, 0.2A) formed by the temperature and current changes and voltage values ​​constitutes the thermal behavior sequence during fast charging. This sequence helps analyze the battery's thermal management state during fast charging.

[0027] Please see Figure 3 The specific steps of S2 are as follows: S201: Based on the fast charging thermal behavior sequence, the sampling point information is extracted in the main control chip according to the time index, the temperature change, voltage value and current change of the corresponding time node are called, and the predicted temperature value of the time point is generated by combining the thermal estimation module of the control chip. The predicted temperature value is then arranged in time order to generate a predicted temperature value sequence. First, based on the fast-charging thermal behavior sequence, the main control chip extracts the temperature change, voltage value, and current change at each time point according to the time index of the sampling points. These data reflect the battery's working state during the charging process. After obtaining this information, the control chip performs calculations through its built-in thermal estimation module. The thermal estimation module predicts the temperature value at the current time point based on a certain algorithm model. This model may involve physical characteristics of the battery such as thermal conduction and thermal radiation. Assuming that at a certain moment, the temperature change is 1℃, the current change is 0.2A, and the voltage value is 4.2V, the chip predicts the battery temperature at that moment to be 31℃ by combining these data with the model. This predicted temperature is then calculated and arranged in chronological order with the predicted temperatures at other time points to form a predicted temperature value sequence. Each item in the sequence corresponds to the predicted temperature value at different time points, providing basic data for subsequent temperature error analysis and processing.

[0028] S202: Based on the predicted temperature value sequence, call the actual temperature value at the corresponding time point, calculate the error value between the predicted temperature value and the actual temperature value, compare the error value with the set thermal safety error range, mark the corresponding range level label according to the judgment result, and obtain the temperature error range level sequence. First, the main control chip extracts the predicted temperature value for each time point from the predicted temperature value sequence and compares it with the actual temperature value. For example, if the predicted temperature is 31℃ and the actual measured temperature is 30.8℃, the temperature error is 0.2℃. This error value is compared with a pre-set thermal safety error range. This error range can be set using experimental data or battery safety operating parameters. For instance, an error range of ±0.5℃ is acceptable, while errors exceeding this range require processing. By comparing these error values ​​with the set range, the main control chip assigns a corresponding range level label to each error value. For example, an error value of 0.2℃ falling within the ±0.5℃ range is assigned a normal level; if the error exceeds this range, it is assigned an abnormal level. This results in a temperature error range level sequence, which shows whether the temperature error at each time point is within the safe range and provides a basis for subsequent error trend analysis.

[0029] S203: Call the level labels of time nodes in the temperature error interval level sequence, and compare the differences with the direction of temperature change within the same time point to determine the error change trend. Then, pair the trend value with time and merge it into the interval level sequence to generate a thermal prediction error level set. First, the control chip extracts the error level label for each time point from the temperature error range level sequence and compares the differences with the direction of temperature change at the same time point. For example, if the temperature change is 1℃ and the error value is 0.2℃, the positive error value indicates an upward trend in temperature. Next, the chip analyzes the trend of error change based on the direction of temperature change (increasing or decreasing). If the direction of error change is consistent with the direction of temperature change, it indicates a potential trend of temperature prediction error accumulation. If the direction of error change is opposite to the direction of temperature change, it indicates that the predicted temperature value may have a certain calibration error. After analysis, the chip pairs these error trends with time, merging the trend information of each time point with its corresponding error level to generate a new thermal prediction error hierarchy. This hierarchy displays the error level and error change trend at each time point, providing more accurate prediction data for the thermal management strategy of the battery management system.

[0030] Please see Figure 4 The specific steps of S3 are as follows: S301: Call the temperature error range level and current change of the time node in the thermal prediction error classification set, calculate the current correction base value at the time point according to the adjustment parameter corresponding to the error level, and generate the current correction base value sequence. First, during execution, the main control chip extracts the temperature error range level for each time node from the thermal prediction error classification set. Assuming that at a certain moment, the error level is normal, the temperature error is 0.2℃, and the current change is 0.5A, the chip calls the corresponding adjustment parameter according to the error level. For example, if the current correction parameter is set to 0.1A according to the normal level, the correction base value is calculated based on this parameter and the current change. If the current change is 0.5A, then the current correction base value is 0.5A + 0.1A = 0.6A. Based on this correction base value, the chip generates a current correction base value sequence. This sequence changes according to the error level and current change at different time nodes, ultimately generating a corresponding current correction amount for each moment, which is used for current adjustment during subsequent charging.

[0031] S302: Based on the temperature error range level and temperature change amount of the time nodes in the thermal prediction error classification set, and combined with the cooling parameters corresponding to the level, calculate the basic value of cooling correction at the time point and generate a sequence of basic values ​​of cooling correction amount. The main control chip extracts the error level and temperature change at each time point from the thermal prediction error classification set. Assuming the temperature error level is normal at a certain moment and the temperature change is 1℃, the chip calls the cooling correction parameter according to the temperature error range level. If the cooling parameter corresponding to the normal level is 0.2℃, then the basic value of the cooling correction is the current temperature change of 1℃ + 0.2℃ = 1.2℃. In this way, the main control chip can calculate the cooling correction amount based on the temperature change and error level at different time points, thus forming a basic value sequence of cooling correction amounts. Assuming the temperature change is 1℃ at that moment and the error is normal, the cooling correction amount is 1.2℃. This correction amount reflects the cooling amount required based on the current state.

[0032] S303: Call the current correction base value sequence and the cooling correction base value sequence, combine them with the corresponding current adjustable amplitude limit and cooling power adjustable amplitude limit, determine whether the base value exceeds the set amplitude, and arrange them in time order to form an adjustment set, generating the charging offset combination value; The main control chip will simultaneously use the current correction base value sequence and the cooling correction base value sequence. Assuming that the correction value in the current correction base value sequence is 0.6A and the cooling correction base value is 1.2℃ at a certain moment, the chip will then compare whether the current correction value and the cooling correction value exceed the preset upper limit of the adjustable current amplitude and the upper limit of the adjustable cooling power amplitude. Assuming that the upper limit of the adjustable current amplitude is 1A and the upper limit of the adjustable cooling power amplitude is 2℃, at this moment, the current correction value of 0.6A does not exceed 1A and the cooling correction value of 1.2℃ does not exceed 2℃, so both are within the set range. The chip will arrange the current correction value and the cooling correction value in chronological order and combine them to form a charging-cooling offset combination value. Finally, it will generate a corresponding charging-cooling offset combination value for each time point for use in the subsequent current and cooling adjustments during the charging process.

[0033] Please see Figure 5 The specific steps of S4 are as follows: S401: Based on the combined value of charging and cooling offsets, match the current offset and cooling offset parameters for the time period in the battery management structure, extract the corresponding adjustment fields in sequence, and write the fields and offset combination values ​​into a preset field sequence to generate the current correction setting sequence and the cooling correction setting sequence. First, the main control chip extracts the current offset and cooling offset parameters for the corresponding time period based on the combined charging and cooling offset values. Assuming the combined charging and cooling offset value for a certain time period is (0.6A, 1.2℃), the chip searches for the corresponding current offset and cooling offset adjustment fields based on the preset battery management structure and offset combination. If the current offset parameter is 0.1A and the cooling offset parameter is 0.2℃, the chip matches the fields with the combined charging and cooling offset values ​​and the preset battery management structure to obtain the corresponding adjustment fields. Next, the current offset and cooling offset fields are structurally mapped to the offset combination value (0.6A, 1.2℃). Finally, these matched fields are written into a preset field sequence to generate a current correction setting sequence and a cooling correction setting sequence. These sequences represent the correction values ​​required during charging and cooling, ensuring that the battery management system accurately adjusts the current and cooling during operation to maintain the battery's optimal operating state.

[0034] S402: Call the current correction setting sequence and the cooling correction setting sequence. Based on the rated value of the output current of the DC-DC converter circuit and the rated value of the heat dissipation power of the liquid cooling plate, determine whether the sequence value exceeds the corresponding rated amplitude limit. Perform amplitude compression processing on the sequence value that exceeds the limit to generate a set of correction setting sequences after amplitude limiting. The main control chip retrieves the current correction setting sequence and the cooling correction setting sequence from the battery management system. Assuming the correction value in the current correction setting sequence is 1.2A at a certain moment and the correction value in the cooling correction setting sequence is 2.5℃ at a certain moment, the chip then checks whether the current correction value and the cooling correction value exceed these rated values ​​based on the rated output current of the DC-DC converter circuit and the rated heat dissipation power of the liquid cooling plate. Assuming the rated current is 1.0A and the rated cooling power is 2.0℃, in this case, the 1.2A in the current correction setting sequence exceeds the rated current of 1.0A, and the 2.5℃ in the cooling correction setting sequence also exceeds the rated cooling power of 2.0℃. Therefore, both of these values ​​need to be compressed. The amplitude compression method can be adjusted by simple scaling. Assuming the compression factor is 0.8, the current correction value of 1.2A will be adjusted to 1.0A (1.2 * 0.8), and the cooling correction value of 2.5℃ will be adjusted to 2.0℃ (2.5 * 0.8). After amplitude compression, the generated set of limited correction settings contains current and cooling values ​​that do not exceed the rated amplitude, ensuring that the equipment will not be overloaded or overheated during operation.

[0035] S403: Based on the set of correction settings after amplitude limiting, extract sequence values ​​according to the field definition specifications of the vehicle internal communication link, assemble the processed field contents into a multi-frame communication structure in sequence, and write the number and format according to the frame structure settings to generate a set of correction control settings frames. The main control chip extracts the corresponding sequence values ​​from the set of correction settings after amplitude compression. Assuming the current correction value after amplitude compression is 1.0A and the cooling correction value is 2.0℃, the chip extracts these sequence values ​​according to the field definition specifications of the vehicle's internal communication link and processes the extracted field content. Specifically, the current correction value and the cooling correction value are mapped to their respective communication fields. Assuming the field corresponding to the current correction value of 1.0A is "current correction field" and the field corresponding to the cooling correction value of 2.0℃ is "cooling correction field", after extracting and matching these fields, the chip assembles them into a multi-frame communication structure in sequence. Each frame structure is identified by a predetermined number and formatted according to the vehicle communication protocol. Finally, the processed communication data is written into the corresponding communication frame set, thereby generating a correction control setting frame set, which is used as the control information related to current and cooling in the vehicle's battery management system for transmission.

[0036] Please see Figure 6 The specific steps of S5 are as follows: S501: Based on the modified control setting frame set, the output current of the DC-DC converter circuit, the heat dissipation power of the liquid cooling plate, and the temperature value of the battery module are recorded sequentially in the battery management structure, and the signals are sampled periodically and organized into a set of periodic operating state values. After receiving the correction control setting frame set, the main control chip periodically collects current, cooling power, and temperature signals according to a preset sampling period (e.g., once per second). Assuming that in a certain period, the current signal is 1.2A, the cooling power is 2.4W, and the temperature is 35°C, the chip will record these signals and update them to the periodic operating state value set. As the sampling period continues, the chip continuously collects these signal data, eventually forming a set encompassing multiple parameters such as current, cooling power, and temperature for subsequent processing and analysis. These sampling period signal data will reflect the actual operating state of the battery management system.

[0037] S502: Based on the current and cooling power signals in the set of periodic operating status values, calculate the current deviation and cooling deviation, and average the deviation values ​​in each period to obtain the current execution deviation value and cooling execution deviation value. The main control chip extracts current and cooling power signals from the set of periodic operating state values ​​and compares them with preset target values. For example, if the current signal in a certain period is 1.2A, the target current is 1.0A, the cooling power is 2.4W, and the target cooling power is 2.0W, then the current deviation is 0.2A, and the cooling deviation is 0.4W. Next, the main control chip averages the deviation values ​​over multiple periods. Assuming the collected current deviation values ​​for multiple periods are 0.2A, 0.3A, 0.1A, etc., and the cooling deviation values ​​are 0.4W, 0.5W, 0.3W, etc., the average of these deviation values ​​yields the current execution deviation value and the cooling execution deviation value. In this way, by calculating the deviation for each period and averaging it, a more accurate execution deviation value can be obtained, providing a basis for subsequent adjustments.

[0038] S503: Combines the current execution deviation value, the cooling execution deviation value, and the temperature data in the set of cycle operation status values, and pairs the deviation with the temperature value in chronological order to generate a set of execution deviation change records; The main control chip matches the calculated current and cooling deviation values ​​with the temperature data in the set of cycle operating state values, ensuring a one-to-one pairing of the deviation value with the corresponding temperature data for each cycle. For example, in one cycle, the current deviation might be 0.2A and the temperature 36°C; in another cycle, the current deviation might be 0.3A and the temperature 37°C. In this way, the changes in current and temperature are recorded chronologically, forming a set of execution deviation change records. This set not only helps to show how the deviation changes over time but also provides data support for analyzing the relationship between current and temperature. These paired data sets will help optimize battery management strategies and ensure the stability of system operation.

[0039] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for fast-charging thermal regulation of lithium-ion batteries using neural network-enhanced model predictive control, characterized in that, Includes the following steps: S1: Acquire the temperature, current, and voltage signals of the battery module and complete the numerical conversion, calculate the temperature and current changes, and combine the results with the sampling time in sequence to form a fast charging thermal behavior sequence; S2: Based on the fast charging thermal behavior sequence, the main control chip calculates the predicted temperature and error, determines the temperature error level according to the preset interval, obtains the error development trend according to the temperature change direction, and generates a thermal prediction error hierarchical set in chronological order. S3: Based on the thermal prediction error hierarchical set, calculate the current and cooling correction base values ​​respectively, compare them with the corresponding upper limit of amplitude, generate the adjustment content set in time order, and obtain the combined value of charging offset; S4: Based on the combined value of charging and cooling offset, a current and cooling correction setting sequence is generated in the battery management structure, and after comparing it with the rated output current and heat dissipation power, the processing sequence is written into the vehicle communication transmission frame to form a correction control setting frame set. S5: Based on the modified control setting frame set, record the actual output current, heat dissipation power and battery temperature, calculate the current and cooling execution deviation, and combine them with the temperature according to the period to form an execution deviation change record set.

2. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 1, characterized in that, The fast charging thermal behavior sequence includes temperature change, current change, sampling timestamp, and time sequence record item number. The thermal prediction error classification set includes predicted temperature value, prediction error value, temperature error interval level, error development trend label, and classification time sequence index. The charging-cooling offset combination value includes current correction base value, cooling correction base value, current adjustable amplitude judgment result, cooling power adjustable amplitude judgment result, and offset time sequence entry. The correction control setting frame set includes current correction setting sequence, cooling correction setting sequence, transmission frame field set, link transmission frame identifier, and frame sequence number time sequence set. The execution deviation change record set includes actual output current record, actual heat dissipation power record, battery module temperature record, current execution deviation value, and cooling execution deviation value.

3. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 1, characterized in that, The specific steps of S1 are as follows: S101: Acquire temperature, current and voltage signals from the lithium-ion battery module, input the signals into the battery management structure, perform numerical conversion according to the sensor sampling frequency, form temperature, current and voltage data arranged in the order of sampling time, and generate a sequence of sampled signal values. S102: Based on the numerical sequence of the sampled signal, calculate the temperature change and current change at adjacent sampling times, construct time difference units by combining time tags, arrange them in time order, and obtain the temperature and current change sequence. S103: Based on the temperature and current change sequence, extract the voltage value, temperature change and current change corresponding to the time node, construct a continuous set including time, voltage and change, and generate a fast charging thermal behavior sequence.

4. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 3, characterized in that, The specific steps of S2 are as follows: S201: Based on the fast charging thermal behavior sequence, extract sampling point information in the main control chip according to the time index, call the temperature change, voltage value and current change of the corresponding time node, combine with the thermal estimation module of the control chip to generate the predicted temperature value of the time point, and arrange them in time order to generate a predicted temperature value sequence. S202: Based on the predicted temperature value sequence, call the actual temperature value at the corresponding time point, calculate the error value between the predicted temperature value and the actual temperature value, compare the error value with the set thermal safety error range, mark the corresponding range level label according to the judgment result, and obtain the temperature error range level sequence. S203: Call the level labels of the time nodes in the temperature error interval level sequence, and compare the differences with the direction of temperature change within the same time point to determine the error change trend. Then, pair the trend value with time and merge it into the interval level sequence to generate a thermal prediction error level set.

5. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 4, characterized in that, The specific steps for S3 are as follows: S301: Call the temperature error interval level and current change of the time node in the thermal prediction error classification set, calculate the current correction base value at the time point according to the adjustment parameter corresponding to the error level, and generate the current correction base value sequence. S302: Based on the temperature error range level and temperature change amount of the time nodes in the heat prediction error classification set, and combined with the cooling parameters corresponding to the level, calculate the basic value of cooling correction at the time point, and generate a sequence of basic values ​​of cooling correction amount. S303: Call the current correction base value sequence and the cooling correction base value sequence, combine them with the corresponding current adjustable amplitude limit and cooling power adjustable amplitude limit, determine whether the base value exceeds the set amplitude, and arrange them in time order to form an adjustment set, generating a charging offset combination value.

6. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 5, characterized in that, The specific steps of S4 are as follows: S401: Based on the combined value of charging and cooling offsets, the current offset and cooling offset parameters of the time period are matched in the battery management structure, the corresponding adjustment fields are extracted in sequence, and the fields are structurally matched with the offset combination values ​​and written into the preset field sequence to generate the current correction setting sequence and the cooling correction setting sequence. S402: Call the current correction setting sequence and cooling correction setting sequence, and determine whether the sequence value exceeds the corresponding rated amplitude limit according to the rated value of the output current of the DC-DC converter circuit and the rated value of the heat dissipation power of the liquid cooling plate. Perform amplitude compression processing on the sequence value that exceeds the limit to generate a set of correction setting sequences after amplitude limiting. S403: Based on the set of correction settings after the amplitude limit, extract the sequence values ​​according to the field definition specifications of the vehicle internal communication link, assemble the processed field contents into a multi-frame communication structure in sequence, and write the number and format according to the frame structure settings to generate a set of correction control settings frames.

7. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 6, characterized in that, The specific steps of S5 are as follows: S501: According to the modified control setting frame set, the output current of the DC-DC converter circuit, the heat dissipation power of the liquid cooling plate and the temperature value of the battery module are recorded sequentially in the battery management structure, and the signals are sampled periodically and organized into a set of periodic operating state values. S502: Based on the current and cooling power signals in the set of periodic operating state values, calculate the current deviation and cooling deviation, and average the deviation values ​​in each period to obtain the current execution deviation value and cooling execution deviation value. S503: Combining the current execution deviation value, the cooling execution deviation value, and the temperature data in the set of periodic operating status values, the deviation and temperature value are paired in chronological order to generate a set of execution deviation change records.

8. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 1, characterized in that, The battery module is a packaged unit composed of multiple lithium-ion cells and integrates temperature, voltage and current sensors; The fast charging thermal behavior sequence is a multi-dimensional time series set composed of temperature changes, current changes, and sampling time obtained by the battery module during fast charging, arranged in order of sampling time.

9. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 1, characterized in that, The main control chip is the core processing unit in the battery management structure. It is an embedded computing chip of MCU, DSP or FPGA, responsible for performing thermal behavior prediction, error judgment and control strategy calculation operations. The predicted temperature and error are the difference between the predicted temperature value calculated by the main control chip and the sampled actual temperature value. The temperature error level is a classification based on the comparison between the predicted error value and the thermal safety error range, indicating the severity of the current temperature deviation. The error development trend is the future evolution trend of the error determined based on the direction of temperature change, which is used to assist the control strategy in determining whether the correction value needs to be adjusted. The thermal prediction error classification set is a set arranged in chronological order, including temperature error interval levels and error development trends; The base value of the current correction is a reference correction value generated by a calculation model based on the temperature error range level and the current change. The base value for cooling correction is a reference correction value generated based on the temperature error range level and the amount of temperature change. The upper limit of amplitude is a set of control target values ​​formed by combining the maximum current adjustment range of the battery system or DC-DC converter circuit, the base value of the current correction amount, the base value of the cooling correction amount, and the upper limit of the adjustable amplitude.

10. The method for fast-charging thermal regulation of lithium-ion batteries using neural network augmentation model predictive control according to claim 1, characterized in that, The current correction setting sequence is a set of time-series setting values ​​generated based on the combined values ​​of the cooling offset, indicating the output current adjustment strategy of the DC-DC converter circuit. The cooling correction setting sequence is a time series set of setting values ​​that indicate the liquid cooling plate cooling power adjustment strategy, generated based on the combined values ​​of the cooling offset. The transmitted frame is a frame format data packet transmitted via the vehicle communication link, including a current and cooling correction setting value field. The correction control setting frame set is a collection of multiple transmission frames arranged in chronological order, used to periodically transmit the currently calculated correction control parameters; The current execution deviation refers to the difference between the actual output current of the DC-DC converter circuit and the set value, which is used to provide feedback on the control effect. The cooling performance deviation refers to the difference between the actual heat dissipation power of the liquid cooling plate and the cooling set value; The execution deviation change record set is a collection of records arranged in chronological order, consisting of periodically collected current execution deviation, cooling execution deviation, and battery module temperature values.

Images