Safety monitoring method for power battery for new energy vehicle

Abstract

The invention relates to a power battery safety monitoring method for a new energy vehicle, and the method comprises the steps: collecting the real-time data of the operation of a battery, calculating the influence of the current temperature change on the internal resistance of the battery, and determining a temperature change adjustment coefficient; in combination with the temperature change adjustment coefficient and the historical operation record, inputting a pre-trained neural network model, and obtaining a current aging degree value; judging according to the current aging degree value and the real-time data in combination with the charge and discharge rate information, if the aging degree value exceeds a preset threshold value, increasing internal resistance compensation, and obtaining a compensated parameter range; extracting reference values of voltage, temperature and current from the compensated parameter range, updating the reference values according to real-time data fluctuation, determining a self-adaptive threshold range, triggering an abnormal signal if the voltage, temperature and current in the self-adaptive threshold range exceed normal fluctuation, and matching a preset response library to obtain a corresponding power limiting instruction; and sending a control signal according to the corresponding power limiting instruction, and adjusting the charging and discharging process to achieve stable operation.

Description

Technical Field

[0001] This invention relates to the field of battery safety technology, and in particular to a method for safety monitoring of power batteries used in new energy vehicles. Background Technology

[0002] As the core of green transportation, the safety of the power batteries in new energy vehicles directly affects the reliability of vehicle operation and the safety of passengers. The charging and discharging process of power batteries involves complex electrochemical reactions, requiring real-time monitoring of parameters such as voltage, temperature, and current to ensure stable system operation. Safety monitoring is not only the technological cornerstone of the development of the new energy vehicle industry, but also crucial for enhancing user trust and promoting market penetration.

[0003] Existing monitoring methods have significant shortcomings in terms of dynamic adaptability and accurate response, making it difficult to meet the safety requirements under complex operating conditions. Most current safety monitoring systems rely on fixed parameter ranges for status judgment, which cannot effectively cope with the diversity and complexity of battery operating environments. For example, the normal range of safety parameters for batteries can vary significantly under different temperatures, aging levels, or high-rate charge / discharge scenarios. Fixed thresholds may lead to misjudging a battery as abnormal in low-temperature environments or failing to identify potential risks in a timely manner under high temperatures and high loads. This inflexible monitoring approach makes it difficult for the system to maintain efficient early warning and response capabilities in changing scenarios. A deeper challenge lies in how to dynamically adjust the threshold range of monitoring parameters based on the real-time operating status of the battery. Battery operating data is affected by various factors, such as ambient temperature, charge / discharge rate, and battery aging level. These factors intertwine and jointly determine the reasonable range of safety thresholds. For example, battery aging leads to increased internal resistance, which in turn affects the normal fluctuation range of voltage and temperature. If the monitoring system cannot dynamically adjust the thresholds according to the aging level, it may miss the early warning opportunity due to the failure to identify abnormal internal resistance in time when an aged battery is operating under high load, leading to potential safety hazards. Therefore, a key issue in the field of safety monitoring of power batteries for new energy vehicles is how to dynamically set threshold ranges for parameters such as voltage, temperature, and current by collecting real-time battery operating data in complex and ever-changing operating environments, combined with factors such as aging degree and ambient temperature, and ensuring that the system triggers precise safety response measures under different levels of anomalies. The core of this problem lies in the monitoring system's ability to adaptively adjust parameter thresholds in dynamic environments and quickly match appropriate response strategies when anomalies occur. For example, in actual operation, a battery pack may experience an abnormal temperature rise during high-temperature fast charging due to increased internal resistance from long-term use. If the system fails to adjust the temperature threshold according to the aging state, it may be unable to limit the charging power or cut off the circuit in time, thereby increasing the risk of overheating or even fire. Solving this problem requires not only technological breakthroughs but also ensuring the robustness and reliability of the monitoring system in multiple scenarios. Summary of the Invention

[0004] The purpose of this invention is to provide a safety monitoring method for power batteries used in new energy vehicles, which improves the adaptability and stability of battery management, extends battery life, and ensures safe and efficient operation under complex working conditions.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] A method for safety monitoring of power batteries for new energy vehicles includes:

[0007] Collect real-time data on the operation of the power battery of the target new energy vehicle, calculate the impact of the current temperature change on the battery internal resistance, and determine the temperature change adjustment coefficient;

[0008] By combining the temperature change adjustment coefficient and historical operating records, the data is input into a pre-trained neural network model to obtain the current aging level value.

[0009] The judgment is made based on the current aging level value and real-time data combined with the charge / discharge rate information. If the aging level value exceeds the preset threshold, internal resistance compensation is added to obtain the parameter range after compensation.

[0010] The reference values ​​of voltage, temperature and current are extracted from the compensated parameter range. The reference values ​​are updated according to the real-time data fluctuations using a dynamic adjustment algorithm to determine the adaptive threshold range. If the voltage, temperature and current exceed the normal fluctuations within the adaptive threshold range, an abnormal signal is triggered. The corresponding power limiting command is obtained by matching the abnormal signal with a preset response library.

[0011] According to the corresponding power limit command, a control signal is sent to adjust the charging and discharging process to achieve stable operation, thereby completing the safety monitoring of the target new energy vehicle power battery.

[0012] Optionally, collecting real-time data on the operation of the target new energy vehicle's power battery includes:

[0013] Set the acquisition frequency and collect real-time data on the voltage, temperature, and current of the target new energy vehicle's power battery through sensors;

[0014] Kalman filtering is used to filter noise from the real-time data to obtain smooth real-time data.

[0015] Optionally, the impact of the current temperature change on the battery's internal resistance is calculated, and the temperature change adjustment coefficient is determined, including:

[0016] Battery status information is determined based on the real-time data;

[0017] Environmental factor parameters are calculated based on battery status information and data acquisition time. If the environmental factor parameters exceed the preset threshold, the temperature change value is extracted from the real-time data to obtain the temperature influence weight.

[0018] The influence of temperature change on the battery's internal resistance is calculated by weighting the temperature effect, thus determining the trend of internal resistance change.

[0019] A linear regression algorithm is used to generate a temperature change adjustment coefficient based on the trend of internal resistance change and the temperature change value.

[0020] Optionally, obtaining the current aging level value includes:

[0021] Historical operation records and ambient temperature parameters are extracted, and after data cleaning, an operation time series is generated.

[0022] After extracting features from the running time series and temperature change adjustment coefficient, the data is input into a pre-trained neural network model to estimate the degree of aging and obtain the equipment aging value.

[0023] By comparing the equipment aging level value with the equipment performance indicators, it is determined whether the aging level exceeds the preset threshold, and the current aging level value is output.

[0024] Optionally, the range of parameters obtained after compensation includes:

[0025] Based on the current aging level value and real-time data, a comprehensive operating condition characteristic is generated by using a data fusion method combined with charge / discharge rate information.

[0026] Determine whether internal resistance compensation is triggered based on the comprehensive operating condition characteristics. If triggered, calculate the internal resistance compensation parameters based on the comprehensive operating condition characteristics.

[0027] Adjust the battery model parameters based on the calculated internal resistance compensation parameters to generate the compensated parameter range.

[0028] Optionally, determining the adaptive threshold range includes:

[0029] Based on the compensated parameter range, the reference values ​​of voltage, temperature, and current are calculated using the sliding window method;

[0030] The data fluctuation is calculated based on real-time data, and the fluctuation distribution is determined. If the fluctuation distribution exceeds the preset fluctuation range, the voltage reference value, temperature reference value, and current reference value are updated through an adaptive algorithm to obtain the adjusted reference value.

[0031] Based on the adjusted baseline value, the adaptive threshold range is calculated using the quantile method to determine the upper and lower limits of the threshold.

[0032] Optionally, obtaining the corresponding power limiting command includes:

[0033] Determine whether voltage, temperature, and current data within the adaptive threshold range exceed normal fluctuations, acquire abnormal states, and generate abnormal signals;

[0034] Abnormal signals are formatted through signal encoding to obtain encoded signal data;

[0035] Based on the encoded signal data, the preset response library is queried, and the matching rules are found using a keyword matching algorithm to obtain the corresponding power limiting command.

[0036] The power output of the device is adjusted by power limiting commands, and the adjusted power parameters are generated.

[0037] Operating status data is extracted from the adjusted power parameters, and time series analysis algorithms are used to detect parameter change trends and determine operating stability. If the operating stability is lower than a preset threshold, an optimization signal is generated, the matching rules in the response library are adjusted, and updated power limiting instructions are obtained.

[0038] Optionally, a control signal is sent according to the corresponding power limit command to adjust the charging and discharging process to achieve stable operation, including:

[0039] According to the corresponding power limit command, control signals are sent to modify the parameters of the charging and discharging process, so as to obtain a matching and precise operating state;

[0040] Acquire real-time operating status data, analyze the deviation between the status data and the precise response requirements, determine whether a stable operating state has been reached, and if the deviation exceeds the preset deviation, use the support vector machine algorithm to classify the operating status data, determine the parameters that need further adjustment, and generate new control signals based on the classification results to update the charging and discharging process and obtain an optimized operating state.

[0041] By monitoring the operational status in real time, obtaining updated status data, verifying whether the requirements for accurate response are met, and obtaining the final stable operational status.

[0042] The beneficial effects of this invention are as follows:

[0043] This invention addresses the internal resistance fluctuations and performance degradation of power batteries under complex operating conditions caused by temperature changes, aging, and charge / discharge rates. It proposes a comprehensive solution: Real-time data such as voltage, temperature, and current are collected by sensors. Kalman filtering effectively removes noise, and an environmental factor model is used to calculate the impact of temperature changes on internal resistance, determining an adjustment coefficient. Based on this coefficient and historical operating records, a neural network is trained to accurately estimate the battery aging level. When the aging level exceeds a threshold, charge / discharge rate information is integrated for internal resistance compensation, dynamically adjusting voltage, temperature, and current baseline values ​​to form an adaptive threshold range. If data fluctuations exceed the range, an abnormal signal is triggered and matched against a preset response library to generate a power limiting command. This, in turn, adjusts the charge / discharge process through control signals to achieve stable operation. This invention significantly improves the adaptability and stability of battery management, extends battery life, and ensures safe and efficient operation under complex conditions. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the 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.

[0045] Figure 1 This is a flowchart of a safety monitoring method for power batteries used in new energy vehicles according to an embodiment of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0048] This embodiment provides a method for safety monitoring of power batteries used in new energy vehicles, such as... Figure 1 As shown, it includes:

[0049] Collect real-time data on the operation of the power battery of the target new energy vehicle, calculate the impact of the current temperature change on the battery internal resistance, and determine the temperature change adjustment coefficient;

[0050] By combining the temperature change adjustment coefficient and historical operating records, the data is input into a pre-trained neural network model to obtain the current aging level value.

[0051] The judgment is made based on the current aging level value and real-time data combined with the charge / discharge rate information. If the aging level value exceeds the preset threshold, internal resistance compensation is added to obtain the parameter range after compensation.

[0052] The reference values ​​of voltage, temperature and current are extracted from the compensated parameter range. The reference values ​​are updated according to the real-time data fluctuations using a dynamic adjustment algorithm to determine the adaptive threshold range. If the voltage, temperature and current exceed the normal fluctuations within the adaptive threshold range, an abnormal signal is triggered. The corresponding power limiting command is obtained by matching the abnormal signal with a preset response library.

[0053] According to the corresponding power limit command, a control signal is sent to adjust the charging and discharging process to achieve stable operation, thereby completing the safety monitoring of the target new energy vehicle power battery.

[0054] Specifically, this embodiment collects real-time data such as voltage, temperature, and current through sensors, effectively removes noise using Kalman filtering, and calculates the impact of temperature changes on internal resistance using an environmental factor model to determine an adjustment coefficient. Based on this coefficient and historical operating records, a neural network is trained to accurately estimate the battery aging level. When the aging level exceeds a threshold, internal resistance compensation is performed by integrating charge / discharge rate information, dynamically adjusting the voltage, temperature, and current reference values ​​to form an adaptive threshold range. If data fluctuations exceed the range, an abnormal signal is triggered and matched against a preset response library to generate a power limiting command. This, in turn, adjusts the charging and discharging process through control signals to achieve stable operation. This embodiment significantly improves the adaptability and stability of battery management, extends battery life, and ensures safe and efficient operation under complex conditions.

[0055] Furthermore, the collection of real-time data on the operation of the target new energy vehicle's power battery includes:

[0056] Set the acquisition frequency and collect real-time data on the voltage, temperature, and current of the target new energy vehicle's power battery through sensors;

[0057] Kalman filtering is used to filter noise from the real-time data to obtain smooth real-time data.

[0058] Specifically, in this embodiment, real-time data such as voltage, temperature and current are collected from the power battery through sensors. Assuming a lithium-ion battery is used, the sensor collects data at a frequency of 100Hz, with a voltage range of 3.0V to 4.2V, a temperature range of 20°C to 60°C, and a current range of -50A to 50A.

[0059] For example, a set of raw data was acquired: voltage 3.8V, temperature 30.5°C, and current 20.3A. However, the data contained Gaussian white noise, with standard deviations of 0.05V for voltage, 0.5°C for temperature, and 0.3A for current. To filter the noise, a Kalman filter algorithm was used. The raw voltage measurement of 3.8V was reduced to 3.79V after Kalman filtering, with the deviation reduced to 0.01V. The temperature was filtered from 30.5°C to 30.4°C, and the current from 20.3A to 20.2A. The smoothness of the filtered data was improved, and the standard deviations decreased to 0.02V, 0.2°C, and 0.1A. Analysis shows that Kalman filtering effectively suppresses noise while preserving the true trend of the signal.

[0060] Furthermore, the impact of the current temperature change on the battery's internal resistance is calculated, and the temperature change adjustment coefficient is determined, including:

[0061] Battery status information is determined based on the real-time data;

[0062] Environmental factor parameters are calculated based on battery status information and data acquisition time. If the environmental factor parameters exceed the preset threshold, the temperature change value is extracted from the real-time data to obtain the temperature influence weight.

[0063] The influence of temperature change on the battery's internal resistance is calculated by weighting the temperature effect, thus determining the trend of internal resistance change.

[0064] A linear regression algorithm is used to generate a temperature change adjustment coefficient based on the trend of internal resistance change and the temperature change value.

[0065] Specifically, this embodiment calculates environmental factor parameters using a pre-defined environmental factor model. This model is based on the Arrhenius equation. By adjusting the filter voltage and current using this model, the effective voltage and current after environmental correction are obtained, forming a continuous data chain. Based on this, the impact of current temperature changes on battery internal resistance is calculated. First, the real-time temperature is monitored as it rises from an initial 25°C to 32°C, a change of ΔT = 7°C. Empirical formula analysis shows an increase in internal resistance of 5.6%, which is verified by Ohm's law V = IR. The filtered voltage of 12.16V divided by the current of 1.91A yields an internal resistance of 0.0636Ω, with the deviation from the model prediction controlled within 8%. This demonstrates the physical mechanism by which increased temperature leads to increased electrolyte viscosity and decreased ion mobility, thereby quantifying the negative impact of thermal effects on battery performance. Finally, the temperature change adjustment coefficient κ was determined. κ was fitted to the historical temperature-internal resistance dataset using a linear regression algorithm. Combined with the current ΔT=7℃, the adjustment coefficient was calculated to be 0.0035. This coefficient is used to compensate for the charging and discharging strategies in real-time battery management. For example, κ can be applied to the SOC estimation algorithm to correct the internal resistance term, ensuring that the output power remains stable above 95%.

[0066] Furthermore, obtaining the current aging level value includes:

[0067] Historical operation records and ambient temperature parameters are extracted, and after data cleaning, an operation time series is generated.

[0068] After extracting features from the running time series and temperature change adjustment coefficient, the data is input into a pre-trained neural network model to estimate the degree of aging and obtain the equipment aging value.

[0069] By comparing the equipment aging level value with the equipment performance indicators, it is determined whether the aging level exceeds the preset threshold, and the current aging level value is output.

[0070] Specifically, this embodiment extracts historical operation records, selecting log data with a cumulative runtime of 4500 hours over the past 365 days. A time series analysis algorithm, such as the ARIMA model, is used to fit the trend, predicting a short-term aging increment of 0.05 units to avoid data noise interference. The adjustment coefficient and historical records are fused as input feature vectors and input to a pre-trained BP neural network model. This model contains 12 neurons in the input layer, two hidden layers each with 20 ReLU activation function neurons, and one sigmoid function neuron in the output layer. The training dataset comes from 1000 similar device samples. The Adam optimizer is used for 500 iterations with a learning rate of 0.001, and the loss function MSE converges to 0.012. Model parameters are updated through backpropagation to minimize prediction error. Finally, the neural network outputs a current aging level value of 0.73, indicating that the device is in a moderate aging stage.

[0071] Furthermore, the range of parameters obtained after compensation includes:

[0072] Based on the current aging level value and real-time data, a comprehensive operating condition characteristic is generated by using a data fusion method combined with charge / discharge rate information.

[0073] Determine whether internal resistance compensation is triggered based on the comprehensive operating condition characteristics. If triggered, calculate the internal resistance compensation parameters based on the comprehensive operating condition characteristics.

[0074] Adjust the battery model parameters based on the calculated internal resistance compensation parameters to generate the compensated parameter range.

[0075] Specifically, this embodiment, based on the current aging level value and real-time data, integrates charge / discharge rate information under complex operating conditions to extract the C-rate. For example, under high temperature and high load conditions, the charging C-rate is 1.5C and the discharging C-rate is 2C. These are then fused using a weighted average algorithm to obtain a comprehensive rate R_f = 0.6 * 1.5 + 0.4 * 2 = 1.7C. This fusion considers operating condition weights to reflect the actual stress impact. Next, it determines whether the aging level value exceeds a preset threshold of 0.8. If SOH = 0.85 > 0.8, an alarm is triggered and compensation logic is entered; otherwise, the original parameters are maintained. This judgment is based on a threshold model to avoid the risk of excessive aging. Finally, the increased internal resistance compensation is calculated using an ohmic model, and the battery model parameters are adjusted to generate the compensated parameter range.

[0076] Further, determining the adaptive threshold range includes:

[0077] Based on the compensated parameter range, the reference values ​​of voltage, temperature, and current are calculated using the sliding window method;

[0078] The data fluctuation is calculated based on real-time data, and the fluctuation distribution is determined. If the fluctuation distribution exceeds the preset fluctuation range, the voltage reference value, temperature reference value, and current reference value are updated through an adaptive algorithm to obtain the adjusted reference value.

[0079] Based on the adjusted baseline value, the adaptive threshold range is calculated using the quantile method to determine the upper and lower limits of the threshold.

[0080] Specifically, this embodiment extracts baseline values ​​for voltage, temperature, and current, assuming a voltage of 10.5V, a temperature of 25°C, and a current of 2.0A. A mean filtering algorithm is used to select the most recent 10 sampled data points and calculate the average value as the baseline. For example, the voltage sample [10.4, 10.6, 10.5, 10.3, 10.7, 10.5, 10.4, 10.6, 10.5, 10.5] has a mean of 10.5V. Temperature and current are processed similarly, resulting in baseline values ​​of 25.2°C and 2.01A, respectively. The dynamic adjustment algorithm updates the baseline values ​​based on real-time data fluctuations, using a sliding window (window size 10) and a weighted moving average algorithm. The weights decrease linearly over time, with the latest data having a weight of 0.3 and the oldest having a weight of 0.05. A new baseline value is calculated; for example, a new voltage sample of 10.8V is added, the oldest sample of 10.4V is removed, and the weighted average is recalculated to obtain 10.52V. Temperature and current are updated in the same way. Determine the adaptive threshold range by using the standard deviation method to calculate the standard deviation of the most recent 10 data points. For example, if the standard deviation of voltage is 0.12V, set the threshold range to the baseline value ± 2 times the standard deviation, i.e., 10.52 ± 0.24V (10.28V to 10.76V). The temperature and current thresholds are calculated similarly, resulting in a temperature range of 23.8°C to 26.6°C and a current range of 1.95A to 2.07A.

[0081] Furthermore, obtaining the corresponding power limiting command includes:

[0082] Determine whether voltage, temperature, and current data within the adaptive threshold range exceed normal fluctuations, acquire abnormal states, and generate abnormal signals;

[0083] Abnormal signals are formatted through signal encoding to obtain encoded signal data;

[0084] Based on the encoded signal data, the preset response library is queried, and the matching rules are found using a keyword matching algorithm to obtain the corresponding power limiting command.

[0085] The power output of the device is adjusted by power limiting commands, and the adjusted power parameters are generated.

[0086] Operating status data is extracted from the adjusted power parameters, and time series analysis algorithms are used to detect parameter change trends and determine operating stability. If the operating stability is lower than a preset threshold, an optimization signal is generated, the matching rules in the response library are adjusted, and updated power limiting instructions are obtained.

[0087] Specifically, this embodiment monitors voltage, temperature, and current data in real time. Assuming the voltage range is set to 220V ± 5% (209V to 231V), the temperature range to -20℃ to 80℃, and the current range to 0A to 10A, with a sampling frequency of 10 times per second, a sliding window algorithm is used. The window size is 60 seconds. The standard deviation of the data within the window is calculated to determine if it exceeds the normal fluctuation range. For example, if the voltage standard deviation exceeds 2V, the temperature standard deviation exceeds 5℃, or the current standard deviation exceeds 0.5A, an abnormal signal is triggered. Data acquisition is transmitted to the embedded controller via the I2C protocol through the sensor. The controller performs real-time operation, executing the sliding window algorithm. If a voltage of 230V, a temperature of 85℃, and a current of 11A are detected, the calculated voltage standard deviation is 1.8V, the temperature standard deviation is 6.2℃, and the current standard deviation is 0.7A. Since the temperature and current exceed the thresholds, an abnormal signal is triggered. An abnormal signal is sent to a preset response library via the CAN bus. This library stores the mapping between abnormal modes and power limiting commands; for example, "temperature > 80℃ and current > 10A" corresponds to "power limited to 50%". The matching algorithm uses a decision tree, with the root node representing the abnormal type, branches representing specific parameter ranges, and leaf nodes representing the command. Analyzing the signal, if "temperature > 80℃ and current > 10A" is matched, a power limiting command to 50% is output. This is adjusted to the inverter via a PWM signal, reducing the power from 1000W to 500W. After the command is executed, a log is recorded in the database, including a timestamp, abnormal parameters, and the limiting command, for subsequent analysis. If the database detects three consecutive instances of the same abnormality, an alarm is automatically triggered and sent to a remote server via the MQTT protocol, ensuring stability and traceability.

[0088] Furthermore, according to the corresponding power limit command, control signals are sent to adjust the charging and discharging process to achieve stable operation, including:

[0089] According to the corresponding power limit command, control signals are sent to modify the parameters of the charging and discharging process, so as to obtain a matching and precise operating state;

[0090] Acquire real-time operating status data, analyze the deviation between the status data and the precise response requirements, determine whether a stable operating state has been reached, and if the deviation exceeds the preset deviation, use the support vector machine algorithm to classify the operating status data, determine the parameters that need further adjustment, and generate new control signals based on the classification results to update the charging and discharging process and obtain an optimized operating state.

[0091] By monitoring the operational status in real time, obtaining updated status data, verifying whether the requirements for accurate response are met, and obtaining the final stable operational status.

[0092] Specifically, this embodiment receives external commands, such as a maximum charging power limit of 50kW and a maximum discharging power of 40kW. The commands are parsed via the CAN bus at 500ms intervals, extracting the power values ​​and storing them in the controller's memory. A proportional-integral (PI) control algorithm is used to adjust the charging and discharging current, setting the proportional coefficient Kp to 0.8 and the integral coefficient Ki to 0.05. The target current is calculated, and the control output is calculated based on the deviation between the actual current and the target current. A PWM signal is generated to adjust the duty cycle of the DC-DC converter, ensuring the current remains stable within the range of 125A±2A. To match the requirement for precise response, battery status data, including voltage, current, and SOC (State of Charge), is collected every 100ms. SOC changes are predicted using a Kalman filter algorithm with a filter gain set to 0.1, and the prediction error is controlled within ±1%. If the SOC approaches 90%, the charging power is automatically reduced to 30kW to avoid overcharging. The analysis process uses historical data for regression to assess battery temperature changes after power adjustment. A safety threshold of 55℃ is set. If the temperature exceeds 50℃, the power is reduced by 10kW and the cooling fan speed is increased to 3000rpm. Stable operation is achieved by real-time monitoring of power fluctuations, with the fluctuation rate controlled within ±5%. If the fluctuation exceeds the limit, a feedback mechanism is triggered to recalculate the PI parameters and update the control signal.

[0093] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made by those skilled in the art to the technical solutions of the present invention without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for safety monitoring of power batteries for new energy vehicles, characterized in that, include: Collect real-time data on the operation of the power battery of the target new energy vehicle, calculate the impact of the current temperature change on the battery internal resistance, and determine the temperature change adjustment coefficient; By combining the temperature change adjustment coefficient and historical operating records, the data is input into a pre-trained neural network model to obtain the current aging level value. The judgment is made based on the current aging level value and real-time data combined with the charge / discharge rate information. If the aging level value exceeds the preset threshold, internal resistance compensation is added to obtain the parameter range after compensation. The reference values ​​of voltage, temperature and current are extracted from the compensated parameter range. The reference values ​​are updated according to the real-time data fluctuations using a dynamic adjustment algorithm to determine the adaptive threshold range. If the voltage, temperature and current exceed the normal fluctuations within the adaptive threshold range, an abnormal signal is triggered. The corresponding power limiting command is obtained by matching the abnormal signal with a preset response library. According to the corresponding power limit command, a control signal is sent to adjust the charging and discharging process to achieve stable operation, thereby completing the safety monitoring of the target new energy vehicle power battery.

2. The method for safety monitoring of power batteries for new energy vehicles according to claim 1, characterized in that, The real-time data collected on the operation of the target new energy vehicle's power battery includes: Set the acquisition frequency and collect real-time data on the voltage, temperature, and current of the target new energy vehicle's power battery through sensors; Kalman filtering is used to filter noise from the real-time data to obtain smooth real-time data.

3. The method for safety monitoring of power batteries for new energy vehicles according to claim 1, characterized in that, Calculate the impact of current temperature changes on battery internal resistance, and determine the temperature change adjustment coefficient, including: Battery status information is determined based on the real-time data; Environmental factor parameters are calculated based on battery status information and data acquisition time. If the environmental factor parameters exceed the preset threshold, the temperature change value is extracted from the real-time data to obtain the temperature influence weight. The influence of temperature change on the battery's internal resistance is calculated by weighting the temperature effect, thus determining the trend of internal resistance change. A linear regression algorithm is used to generate a temperature change adjustment coefficient based on the trend of internal resistance change and the temperature change value.

4. The method for safety monitoring of power batteries for new energy vehicles according to claim 1, characterized in that, Obtaining the current aging level value includes: Historical operation records and ambient temperature parameters are extracted, and after data cleaning, an operation time series is generated. After extracting features from the running time series and temperature change adjustment coefficient, the data is input into a pre-trained neural network model to estimate the degree of aging and obtain the equipment aging value. By comparing the equipment aging level value with the equipment performance indicators, it is determined whether the aging level exceeds the preset threshold, and the current aging level value is output.

5. The method for safety monitoring of power batteries for new energy vehicles according to claim 1, characterized in that, The range of parameters obtained after compensation includes: Based on the current aging level value and real-time data, a comprehensive operating condition characteristic is generated by using a data fusion method combined with charge / discharge rate information. Determine whether internal resistance compensation is triggered based on the comprehensive operating condition characteristics. If triggered, calculate the internal resistance compensation parameters based on the comprehensive operating condition characteristics. Adjust the battery model parameters based on the calculated internal resistance compensation parameters to generate the compensated parameter range.

6. The safety monitoring method for power batteries used in new energy vehicles according to claim 1, characterized in that, Determining the adaptive threshold range includes: Based on the compensated parameter range, the reference values ​​of voltage, temperature, and current are calculated using the sliding window method; The data fluctuation is calculated based on real-time data, and the fluctuation distribution is determined. If the fluctuation distribution exceeds the preset fluctuation range, the voltage reference value, temperature reference value, and current reference value are updated through an adaptive algorithm to obtain the adjusted reference value. Based on the adjusted baseline value, the adaptive threshold range is calculated using the quantile method to determine the upper and lower limits of the threshold.

7. The safety monitoring method for power batteries in new energy vehicles according to claim 1, characterized in that, Obtaining the corresponding power limit command includes: Determine whether voltage, temperature, and current data within the adaptive threshold range exceed normal fluctuations, acquire abnormal states, and generate abnormal signals; Abnormal signals are formatted through signal encoding to obtain encoded signal data; Based on the encoded signal data, the preset response library is queried, and the matching rules are found using a keyword matching algorithm to obtain the corresponding power limiting command. The power output of the device is adjusted by power limiting commands, and the adjusted power parameters are generated. Operating status data is extracted from the adjusted power parameters, and time series analysis algorithms are used to detect parameter change trends and determine operating stability. If the operating stability is lower than a preset threshold, an optimization signal is generated, the matching rules in the response library are adjusted, and updated power limiting instructions are obtained.

8. The safety monitoring method for power batteries used in new energy vehicles according to claim 1, characterized in that, According to the corresponding power limit command, control signals are sent to adjust the charging and discharging process to achieve stable operation, including: According to the corresponding power limit command, control signals are sent to modify the parameters of the charging and discharging process, so as to obtain a matching and precise operating state; Acquire real-time operating status data, analyze the deviation between the status data and the precise response requirements, determine whether a stable operating state has been reached, and if the deviation exceeds the preset deviation, use the support vector machine algorithm to classify the operating status data, determine the parameters that need further adjustment, and generate new control signals based on the classification results to update the charging and discharging process and obtain an optimized operating state. By monitoring the operational status in real time, obtaining updated status data, verifying whether the requirements for accurate response are met, and obtaining the final stable operational status.

Images