High-weather-resistance wide-temperature storage battery monitoring method and system based on power transmission line on-line monitoring device, terminal equipment and storage medium

Abstract

The invention relates to the technical field of wide-temperature battery monitoring and management, in particular to a high-weather-resistance wide-temperature storage battery monitoring method and system based on a power transmission line online monitoring device, terminal equipment and a storage medium, and the method comprises the following steps: S1, collecting voltage, current, power and environment data in real time; s2, constructing a time sequence data matrix of the monitoring device, the environment and the battery characteristics; s3, dynamically adjusting a threshold value according to the joint entropy and the historical state; s4, fine wide-temperature battery feedback analysis is carried out; and S5, if the entropy value continuously increases, executing backtracking analysis. According to the invention, performance monitoring and service life management of the wide-temperature battery can be optimized, and the battery is prevented from being damaged due to excessive discharge or charging at an extreme temperature; the interactive influence of the environment and the monitoring equipment can be accurately evaluated, the accuracy of battery state prediction is remarkably improved, and the overall stability and efficiency of a power transmission line monitoring system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wide-temperature battery monitoring and management, and is a highly weather-resistant wide-temperature storage battery monitoring method, system, terminal equipment and storage medium based on an online monitoring device for a power transmission line. Background Art

[0002] In the field of wide-temperature battery monitoring and management, battery performance monitoring and information feedback optimization are of great significance. Traditional battery monitoring systems mainly rely on a single monitoring indicator or centralized data processing method. However, this method has many limitations. For example, centralized systems cannot flexibly respond to multi-dimensional environmental changes, and under extreme conditions such as high or low temperatures, it is easy to ignore the dynamic impact of the environment on battery life, resulting in increased system instability, which in turn affects the normal operation of the equipment.

[0003] Existing wide-temperature battery monitoring systems usually evaluate the battery status in extreme environments by collecting basic parameters such as battery voltage, current, and temperature. However, most of these systems use fixed time steps and parameter combinations, which cannot fully reflect the performance changes of batteries in complex environments. With the long-term use of batteries and equipment, traditional monitoring systems are prone to problems such as monitoring lag and data redundancy, resulting in the inability to capture early abnormalities of batteries in a timely manner.

[0004] In recent years, information feedback models based on the interaction between environmental characteristics and battery status have gradually been applied to the field of battery management. By constructing an entropy-based feedback model, the data redundancy problem in traditional methods has been partially solved. However, the existing battery monitoring system based on information entropy still has many shortcomings. For example, when dealing with the multi-dimensional interaction between batteries and the environment, the existing models usually use a matrix direct splicing method. This method is computationally complex and lacks optimization of multi-dimensional data, resulting in the monitoring system being unable to operate effectively in a high-dimensional environment. Most of the existing time step adjustment mechanisms are fixed and cannot flexibly respond to the dynamic changes of batteries under different temperature conditions, affecting the real-time and efficiency of the system.

[0005] In addition, traditional wide-temperature battery monitoring systems also have deficiencies in abnormal warning mechanisms. Most of the existing system warning mechanisms rely on single temperature or voltage monitoring, which triggers warnings when certain conditions are abnormal. It is easy to ignore the impact of complex interactions under different environmental conditions on the battery status, resulting in battery performance degradation that cannot be detected in time.

[0006] For example, in the prior art, Satoshi Ohara et al. proposed a battery monitoring system based on the interaction between the environment and the device. The system improves the battery life by real-time monitoring of the battery voltage, current and temperature. However, the system fails to dynamically adjust the time step and fails to optimize the information interaction under high-dimensional data, which affects its application in complex environments.

[0007] Although the existing technology has made certain improvements in battery status monitoring and information feedback optimization, there is still much room for improvement in dealing with complex environments and high-dimensional data interactions, especially in multi-dimensional data fusion and time step dynamic adjustment mechanism. The shortcomings of the existing technology limit the real-time performance and monitoring accuracy of the system. Specifically, the existing technology has the following defects: (1) The efficiency of multi-dimensional data interaction processing is low. The existing wide-temperature battery monitoring system usually uses a matrix direct splicing method to process multi-dimensional data interaction. This method is computationally complex when processing large amounts of data. Especially in extreme environments, the multi-dimensional feature interactions of batteries, environments and monitoring equipment are difficult to capture in a timely manner, resulting in poor real-time performance of the system and inability to effectively deal with high-dimensional data in complex scenarios. (2) The time step adjustment mechanism lacks flexibility. Most existing systems use a fixed time step for data collection and monitoring, ignoring the dynamic changes in battery performance and environmental conditions. This results in the system being unable to adjust the monitoring frequency in a timely manner in scenarios where battery performance fluctuates greatly, resulting in anomalies that cannot be discovered in a timely manner, thereby affecting the battery life and system stability. (3) The early warning mechanism responds laggingly. The current early warning mechanism usually relies on a single or a few parameters (such as temperature, voltage, etc.). When the battery is at extreme temperatures or the load changes dramatically, the system's early warning response is slow and cannot fully capture the battery's state changes in a complex environment, resulting in potential failure risks not being discovered in time, affecting the reliability of the system. (4) The network architecture model has slow feedback. The existing monitoring system fails to fully integrate the dynamic interaction information between the battery and the environment and equipment, and lacks effective real-time computing and processing capabilities. This makes the system have more information redundancy when processing high-dimensional data, resulting in low monitoring efficiency and inability to accurately identify the factors that most significantly affect the battery status, thereby affecting the overall prediction ability of the system. Summary of the invention

[0008] The present invention provides a highly weather-resistant and wide-temperature battery monitoring method, system, terminal equipment and storage medium based on an online monitoring device for a transmission line, which overcomes the deficiencies of the above-mentioned prior art and can effectively solve the problems of the prior art such as low efficiency in multi-dimensional data interactive processing, lack of flexibility in the time step adjustment mechanism, delayed response of the early warning mechanism, and slow feedback of the network architecture model.

[0009] One of the technical solutions of the present invention is achieved by the following measures: A high weather resistance and wide temperature storage battery monitoring method based on an online monitoring device for a transmission line comprises the following steps:

[0010] Step S1, collecting voltage, current, power and environmental data in real time;

[0011] Step S2, constructing a time series data matrix of monitoring device, environment and battery characteristics;

[0012] Step S3, dynamically adjusting the threshold according to the joint entropy and historical status;

[0013] Step S4, performing refined wide temperature battery feedback analysis;

[0014] Step S5: If the entropy value continues to increase, perform backtracking analysis.

[0015] The following is a further optimization and / or improvement of one of the above-mentioned technical solutions:

[0016] In the above step S1, the power requirement of the monitoring device can be set to P device (t), then

[0017] P device-1 (t) = V device (t)·I device (t)

[0018] Where V device (t) and I device (t) are the voltage and current requirements of the monitoring device at time t, respectively;

[0019] Considering the frequency change monitoring device power demand is expressed as:

[0020] P device (t) = P0 + α f(x)

[0021] Where P0 is the power consumption during normal operation, α is the proportional constant between frequency and power consumption, and f(x) is the monitoring frequency of the monitoring device;

[0022] Construct a monitoring device power change rate, which represents the rate of change of the device power demand:

[0023]

[0024] Build environmental conditions, including real-time temperature T(t), temperature gradient The maximum temperature of the day T max and the minimum temperature T min ; Among them, the temperature gradient It represents the rate of change of temperature over time, that is:

[0025]

[0026] The real-time operating parameters of the wide temperature battery include the battery voltage V b (t), current I b (t), remaining charge SOC(t), battery temperature T b (t), real-time battery output power P b(t), battery charge and discharge rate and the internal resistance of the battery;

[0027] Among them, the remaining power SOC(t) is calculated according to the rated capacity and current change of the battery:

[0028]

[0029] In the formula, C r is the rated capacity of the battery, Δt is the time interval;

[0030] Among them, the charge and discharge rate Calculated based on the battery current changes:

[0031]

[0032] In the formula, C r is the rated capacity of the battery.

[0033] In the above step S2, the time series data matrix X of the monitoring device can be constructed device (t), time series data matrix X of environmental conditions env (t) and the battery characteristic matrix X battery (t):

[0034]

[0035] Construct the joint entropy of the wide temperature battery timing joint matrix:

[0036] Among them, each matrix dimension is the monitoring device matrix X device :n×4, environmental condition matrix X env :n×4, battery feature matrix X battery :n×7;

[0037] The battery interacts with the environment information and calculates the battery status X battery and the environmental state X env The information entropy between battery ;X env ), find out the most influential dimensions of environmental characteristics:

[0038]

[0039] In the formula, represents the i-th feature in the environment matrix;

[0040] That is, the information entropy of each environmental dimension is calculated by the above formula, and the dimension with the highest information entropy value is selected as the main factor for weighted calculation;

[0041] Calculate battery status X batteryand monitoring device X device The information entropy between battery ;X device ), identify the dimensions in the monitoring device matrix that have the greatest impact on the battery system:

[0042]

[0043] In the formula, represents the i-th feature in the monitoring device matrix;

[0044] Compare the information entropy values ​​of different features in the monitoring device and select the dimension with the greatest impact;

[0045] Filter out the feature dimension with the largest weight:

[0046] According to the calculated information entropy value, the feature dimensions with the strongest information interaction, that is, the features with the largest information entropy value, are selected from the environment and monitoring device matrices respectively:

[0047]

[0048] Filter the feature dimension with the largest weight from the monitoring device matrix:

[0049]

[0050] Based on the calculation of mutual information, the feature dimension with the largest weight in the environment and monitoring device matrix is ​​selected for the calculation of weighted joint entropy;

[0051] That is, screening after calculating mutual information: sort the mutual information of each dimension and select the feature with the largest mutual information value;

[0052] For the selected feature dimensions, features with higher information entropy will receive greater weights:

[0053]

[0054] in, is the weight of the i-th feature of the environment matrix, normalized based on its information entropy value with the battery;

[0055] is the weight of the i-th feature of the monitoring device matrix, normalized based on its information entropy value with the battery;

[0056] Finally, the joint entropy of the temporal interaction of the three is calculated:

[0057]

[0058] in, is the weight value of the feature dimension with the strongest information interaction selected from the environment matrix, is the weight value of the feature dimension with the strongest information interaction selected from the monitoring device matrix, is the joint probability distribution between battery features, environmental features, and device features, is the joint entropy.

[0059] In the above step S3, the calculation formula for dynamically adjusting the threshold value can be:

[0060]

[0061] In the formula, H threshold (t) is the dynamic threshold at time t, H baseline is the initial baseline entropy value of the system, γ is the adjustment coefficient, which is used to control the sensitivity of the threshold, N is the length of the time window, which means considering the entropy value history of the past N time steps, H weighted (i) is the weighted joint entropy calculated at time step i;

[0062] Current weighted joint entropy H weighted (t) exceeds the dynamic threshold H threshold (t), then based on the subsequent step S4, refined wide temperature battery feedback is performed:

[0063] H weighted (t)≥H threshold (t)

[0064] The above step S4 may specifically include the following steps:

[0065] Step S4.1, screening the battery features based on entropy weights before the recursive network to select the features that contribute most to the prediction of battery life and performance;

[0066] Step S4.2, before feature extraction, a dual-branch recursive network architecture is proposed, which can independently process environmental features and monitoring device features, and interactively model these features with battery features respectively, with each branch focusing on one type of feature to ensure that the model can distinguish and refine the different effects of the environment and load on the battery, and is intended to enhance feature discrimination;

[0067] Step S4.3, dynamically adjust the time step according to the current input features and hidden states through the network, so as to speed up the sampling frequency when the features change drastically, and reduce unnecessary updates when the features change steadily, thereby improving the efficiency and accuracy of the prediction as a whole;

[0068] Step S4.4, in a dual-branch recursive network architecture, the environmental features and the monitoring features are processed as independent inputs for feature extraction;

[0069] Step S4.5, the final output head uses a dual output head for output; by integrating the comprehensive characteristics of the environment and load branch output, the system can predict the entropy value of the battery at the next moment to evaluate the stability of the battery state; the system can predict the remaining life of the battery based on the fused feature output.

[0070] The above step S5 may specifically include the following steps:

[0071] Step S5.1, the system automatically backtracks to check the environmental features that interact most strongly with the battery matrix, especially the temperature, temperature gradient, maximum temperature and minimum temperature. If the ambient temperature is too high or too low, the system backtracks the relationship between the temperature change and the battery health status to identify whether there is a specific temperature that causes the accelerated decay of the battery life;

[0072] Step S5.2, trace back the interaction record between the monitoring device characteristics and the battery to check whether there is a specific load change or power demand that causes a decrease in battery life. Frequent high-load use will cause the battery life to decline rapidly.

[0073] Step S5.3: If the joint entropy value predicted by the system continues to rise, it indicates that the interaction uncertainty between the battery, the environment and the monitoring device increases. At this time, the interaction mode between the three is traced back to check whether the battery is in some extreme environment, resulting in a significant decrease in battery performance;

[0074] Step S5.4, based on the above backtracking results, when the backtracking analysis finds that specific environment or load conditions cause the battery life to decline rapidly, if it is in a high temperature environment, the discharge rate is reduced, if it is in a low temperature environment, the charging efficiency is improved, and the battery preheating is increased; and when it is predicted that the entropy uncertainty continues to increase according to the backtracking results, the monitoring frequency and feature analysis are directly encrypted, the power demand of the transmission line monitoring device is allocated to multiple batteries or different time periods, the impact of a single high load on the battery life is reduced, and the power wall is automatically set.

[0075] The second technical solution of the present invention is achieved by the following measures: a high weather resistance and wide temperature range battery monitoring system based on an online monitoring device for a transmission line, comprising:

[0076] The monitoring environment initialization module is used to initialize the monitoring environment, connect the wide-temperature battery monitoring device, set monitoring parameters, and ensure the normal operation of the device;

[0077] The entropy feedback information flow optimization module is used to dynamically select the feature dimensions with the strongest information interaction based on the monitoring data flow and environmental conditions, avoid redundant data, improve monitoring accuracy, and output the optimized monitoring data flow;

[0078] The real-time monitoring execution module is used to perform full-range monitoring of the wide-temperature battery according to the optimized data stream, and obtain the working parameters of the battery under different ambient temperatures, load conditions and power requirements;

[0079] The data processing module is used to sort and normalize the data obtained by the real-time monitoring execution module, generate a battery feature matrix, an environmental feature matrix and a monitoring device matrix, and input these data into a multidimensional entropy calculation model for feature analysis;

[0080] A joint entropy calculation module is used to calculate the joint entropy value between the battery, the environment and the monitoring device based on the matrix generated by the data processing module, reflect the stability of the battery system and its complexity under different environmental conditions, and generate a real-time risk assessment report based on the joint entropy value;

[0081] A dynamic early warning and feedback module is used to determine whether the battery system is in an unstable state based on the comparison between the joint entropy value and the preset threshold;

[0082] The display module is used to display various real-time data, joint entropy analysis results, risk assessment reports and the operating status of the monitoring device during the battery monitoring process, and provide a visual prediction of the battery health trend.

[0083] The third technical solution of the present invention is achieved through the following measures: a terminal device, including a memory and a processor, the memory stores a program that can be run on the processor, and when the processor executes the program, the above-mentioned high weather resistance and wide temperature battery monitoring method based on the transmission line online monitoring device is implemented.

[0084] The fourth technical solution of the present invention is achieved through the following measures: a storage medium, the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to realize the above-mentioned high weather resistance and wide temperature battery monitoring method based on the transmission line online monitoring device.

[0085] The present invention optimizes the performance monitoring and life management of wide temperature batteries by real-time monitoring of the status data of batteries, environment and monitoring devices, and avoids damage to the batteries due to excessive discharge or charging at extreme temperatures. On the other hand, the system accurately evaluates the interactive effects of the environment and monitoring equipment through a dual-branch recursive network architecture, significantly improving the accuracy of battery status prediction. In addition, the present invention also has the ability to adaptively adjust the monitoring frequency to ensure that the battery life is maximized in a complex environment, and realizes efficient information processing through dynamic adjustment, thereby improving the overall stability and efficiency of the transmission line monitoring system. The present invention proposes a feature screening and weight optimization model based on information entropy, which screens out the most influential feature dimensions by calculating the mutual information value between battery features and environmental and monitoring device features, and avoids the high-dimensional data redundancy problem caused by directly splicing all dimensions. Then, the key features screened out are weighted by the entropy weight optimization model to ensure that the most important dimensions in information interaction are given priority. Finally, the model makes entropy calculation more efficient by reducing redundant information, while improving the accuracy of battery status monitoring. The present invention proposes a learnable time step adjustment mechanism, which dynamically adjusts the time step through a recursive neural network based on the changes in the current input features and hidden states. When the battery performance or environmental conditions change dramatically, the system will shorten the time step and increase the monitoring frequency to capture the abnormal conditions of the battery in time; when the conditions are stable, the system will extend the time step and reduce the monitoring frequency, thereby improving the monitoring efficiency of the system. This dynamic step adjustment enables the system to predict the battery status more efficiently and accurately in a complex environment. The present invention innovatively proposes a dual-branch recursive neural network architecture, which processes environmental features and monitoring equipment features respectively, and independently models them in combination with battery features. Each branch focuses on one type of feature to ensure that the model can accurately capture the different effects of environmental conditions and monitoring equipment on the battery status. In addition, the system introduces a dynamic fusion mechanism in the model output stage, which can adjust the weights of the two branches in real time according to the current input features, thereby improving the accuracy of the model's prediction of battery life and performance. The present invention introduces a wide-temperature battery feedback information flow optimization model. At each time step, the system dynamically selects the most valuable information flow between the battery, environment and monitoring equipment to avoid directly splicing all information for entropy calculation, thereby reducing the computational complexity brought by high-dimensional data. Through this optimization, the system can more flexibly adjust the influence of each feature dimension to ensure that multi-dimensional entropy calculation is more efficient. Finally, a joint entropy evaluation model is used to conduct a global evaluation of the battery's performance under a wide range of temperature conditions to capture any abnormal conditions that may occur in the battery in its early stages. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 It is a schematic diagram of the overall process of a high weather resistance and wide temperature storage battery monitoring method based on an online monitoring device for transmission lines according to an embodiment of the present invention.

[0087] Figure 2 It is a flow chart of step S1 of the high weather resistance and wide temperature range battery monitoring method based on the online monitoring device for transmission lines according to an embodiment of the present invention.

[0088] Figure 3 It is a flow chart of step S2 of the high weather resistance and wide temperature range battery monitoring method based on the online monitoring device for transmission lines according to an embodiment of the present invention.

[0089] Figure 4 It is a flow chart of step S4 of the high weather resistance and wide temperature range battery monitoring method based on the online monitoring device for transmission lines according to an embodiment of the present invention.

[0090] Figure 5 It is a flow chart of step S5 of the high weather resistance and wide temperature range battery monitoring method based on the online monitoring device for transmission lines according to an embodiment of the present invention. DETAILED DESCRIPTION

[0091] The present invention is not limited by the following embodiments, and specific implementation methods can be determined based on the technical solution of the present invention and actual conditions.

[0092] The present invention will be further described below in conjunction with embodiments:

[0093] Embodiment 1: As attached Figure 1 To Attachment Figure 5 As shown, the high weather resistance and wide temperature storage battery monitoring method based on the transmission line online monitoring device comprises the following steps:

[0094] Step S1, real-time collection of voltage, current, power and environmental data. In this embodiment, the monitoring device of the transmission line will run continuously in real time, and then perform high-frequency monitoring of the line's current, voltage, power, environmental parameters, etc. The monitoring device requires a battery to maintain a stable power output. Therefore, in step S1, the power demand of the monitoring device is set to P device (t), depends on the working status of the monitoring equipment;

[0095] P device-1 (t) = V device (t)·I device (t)

[0096] Where V device (t) and I device (t) are the voltage and current requirements of the monitoring device at time t, respectively;

[0097] The relationship between the monitoring frequency and battery consumption of the monitoring device is that under extreme conditions, the monitoring device needs a higher acquisition frequency to obtain more detailed data. For example, under low temperature, high temperature, overload, or failure, the monitoring frequency f(x) of the monitoring device will increase sharply according to actual requirements, resulting in increased battery power consumption. Therefore, considering the frequency change, the power demand of the monitoring device is expressed as:

[0098] P device (t) = P0 + α f(x)

[0099] Where P0 is the power consumption during normal operation, α is the proportional constant between frequency and power consumption, and f(x) is the monitoring frequency of the monitoring device;

[0100] Based on this, the power change rate of the monitoring device is constructed, which represents the rate of change of the power demand of the equipment. This data feature further captures the trend of feedback load changes:

[0101]

[0102] Construct environmental conditions. Environmental characteristics also have a coupled effect on the operation of the monitoring device and the battery. Specific data include: real-time temperature T(t), temperature gradient The maximum temperature of the day T max and the minimum temperature T min ; Among them, the temperature gradient It represents the rate of change of temperature over time, that is:

[0103]

[0104] The real-time operating parameters of the wide temperature battery include the battery voltage V b (t), current I b (t), remaining charge SOC(t), battery temperature T b (t), real-time battery output power P b (t), battery charge and discharge rate and the internal resistance of the battery;

[0105] Among them, the remaining power SOC(t) is calculated according to the rated capacity and current change of the battery:

[0106]

[0107] In the formula, C r is the rated capacity of the battery, Δt is the time interval;

[0108] Among them, the charge and discharge rate Calculated based on the battery current changes:

[0109]

[0110] In the formula, C r is the rated capacity of the battery.

[0111] Step S2: Construct a time series data matrix of monitoring device, environment and battery characteristics. Based on the environment, monitoring device and wide temperature battery parameters collected in step S1, in this embodiment, in step S2, construct a time series data matrix X of the monitoring device device (t), time series data matrix X of environmental conditions env (t) and the battery characteristic matrix X battery (t):

[0112]

[0113] A wide temperature battery timing joint matrix joint entropy is constructed:

[0114] Among them, each matrix dimension is the monitoring device matrix X device :n×4, environmental condition matrix X env :n×4, battery feature matrix X battery :n×7;

[0115] The dimensions of the matrix in the conventional method must be kept consistent during the entropy calculation process. The joint entropy H(X battery ,X env ,X device ) is calculated as follows: These three matrices are concatenated into a new matrix in the column direction, that is, X battery , X env , X device Splice them into an n×15 matrix by column. The specific process is as follows: battery ,X env and X device The concatenation is done along the column direction. The concatenated matrix dimension is n×(7+4+4=15). The joint probability distribution p(x battery ,x env ,x device ), according to the formula of joint entropy: H(x battery ,x env ,x device )=-∑p(x battery ,x env ,x device )logp(x battery ,x env ,x device) calculates the joint entropy to reflect the overall uncertainty of the battery, environment, and device load. However, due to the large number of vertical dimensions, this calculation method will make the calculation too complicated when there are too many points in time, that is, when there are many matrix rows, and the general edge computing performance is difficult to meet.

[0116] Therefore, based on this, the present invention proposes a wide-temperature battery feedback information flow optimization model. Based on the information flow of the battery, environment and equipment, by constructing an information flow model, the most valuable information is dynamically selected at each time step to calculate the entropy with the wide-temperature battery matrix, instead of splicing all the information. That is, the operator is proposed to avoid information redundancy and high-dimensional problems by dynamically adjusting the influence of each dimension:

[0117] The battery interacts with the environment information and calculates the battery status X battery and the environmental state X env The information entropy between battery ;X env ), find out the most influential dimensions of environmental characteristics:

[0118]

[0119] In the formula, represents the i-th feature in the environment matrix;

[0120] That is, the information entropy of each environmental dimension is calculated by the above formula, and the dimension with the highest information entropy value is selected as the main factor for weighted calculation;

[0121] Calculate battery status X battery and monitoring device X device The information entropy between battery ;X device ), identify the dimensions in the monitoring device matrix that have the greatest impact on the battery system:

[0122]

[0123] In the formula, represents the i-th feature in the monitoring device matrix;

[0124] Compare the information entropy values ​​of different features (voltage, current, etc.) in the monitoring device and select the dimension with the greatest impact;

[0125] Filter out the feature dimension with the largest weight:

[0126] According to the calculated information entropy value, the feature dimensions with the strongest information interaction, that is, the features with the largest information entropy value, are selected from the environment and monitoring device matrices respectively:

[0127]

[0128] Filter the feature dimension with the largest weight from the monitoring device matrix:

[0129]

[0130] Based on the calculation of mutual information, the feature dimension with the largest weight in the environment and monitoring device matrix is ​​selected for the calculation of weighted joint entropy;

[0131] That is, screening after calculating mutual information: sort the mutual information of each dimension and select the feature with the largest mutual information value;

[0132] For the selected feature dimensions, features with higher information entropy will receive greater weights:

[0133]

[0134] in, is the weight of the i-th feature of the environment matrix, normalized based on its information entropy value with the battery;

[0135] is the weight of the i-th feature of the monitoring device matrix, normalized based on its information entropy value with the battery;

[0136] Therefore, based on this, in this embodiment, it is only necessary to splice the battery feature matrix (n*7) and the time series data matrix of the monitoring device with the strongest feature dimension of information interaction (n*1), and the time series data matrix of the environmental conditions X env The strongest characteristic dimension of information interaction in (t) (n*1);

[0137] Based on this, the joint entropy of the temporal interaction of the three is finally calculated:

[0138]

[0139] in, is the weight value of the feature dimension with the strongest information interaction selected from the environment matrix, is the weight value of the feature dimension with the strongest information interaction selected from the monitoring device matrix, is the joint probability distribution between battery features, environmental features, and device features, is the joint entropy.

[0140] Based on this, in step S2, an improved wide-temperature battery feedback information flow optimization model is proposed. By screening out the characteristic dimensions with the strongest information interaction, the direct splicing calculation of high-dimensional data is avoided, making the calculation of multi-dimensional entropy more efficient and the final joint entropy calculation more sensitive. Further based on this, the joint entropy calculated by the present invention will be used as a global indicator for system stability evaluation. Through real-time monitoring of multi-dimensional entropy, early abnormal conditions of the battery under wide temperature conditions can be captured; when the entropy value of the battery at certain temperatures increases significantly, it indicates that the uncertainty of the system is increasing, and the early warning mechanism of the following step S3 in the present invention is triggered in advance to prevent premature failure of the battery in high or low temperature environments.

[0141] Step S3, dynamically adjust the threshold according to the joint entropy and historical status. In step S2, the present invention or the joint entropy of the temporal interaction of the three as a whole and the strongest feature of the interaction with the wide temperature battery information in the environment and monitoring device matrix, the following invention proposes a formula for dynamically adjusting the threshold, so that the threshold is adaptively adjusted according to the historical state of the system, rather than a fixed static value, so that it can more flexibly respond to fluctuations and long-term trends in system operation, and decide whether to perform subsequent refined wide temperature battery weather resistance testing. In step S3, the calculation formula for dynamically adjusting the threshold is:

[0142]

[0143] In the formula, H threshold (t) is the dynamic threshold at time t, H baseline is the initial baseline entropy value of the system (determined according to the historical average), γ is the adjustment coefficient, which is used to control the sensitivity of the threshold, N is the length of the time window, which means considering the entropy value history of the past N time steps, H weighted (i) is the weighted joint entropy calculated at time step i;

[0144] Current weighted joint entropy H weighted (t) exceeds the dynamic threshold H threshold (t), then based on the subsequent step S4, refined wide temperature battery feedback is performed:

[0145] H weighted (t)≥H threshold (t)

[0146] Step S4: Perform refined wide-temperature battery feedback analysis.

[0147] In the above steps, the present invention determines whether there is uncertainty in the system as a whole by calculating the joint entropy. When the joint entropy value is high, it indicates that there is greater complexity and information interaction between the battery, environmental conditions and monitoring devices. This high entropy value indicates that the system is in an unstable state, and there may be some external factors that are having a significant impact on the battery performance. However, although a high joint entropy value can indicate instability, it cannot clearly indicate which specific factors cause this instability. In the above steps, the present invention identifies two features with the strongest interaction with the battery through a wide temperature battery feedback information flow optimization model: one each in the environmental conditions and the monitoring device. These two features are the main reasons for the increase in the joint entropy value and the instability of the system. Therefore, based on this, the present invention proposes a bidirectional spatiotemporal prediction model, that is, in summary, the joint entropy provides a global system complexity assessment to help determine whether the system is stable. The two key features provide specific instability factors, which makes the weather resistance analysis of the wide temperature battery more focused, and the bidirectional spatiotemporal prediction model evaluates future instability factors.

[0148] In this embodiment, step S4 specifically includes the following steps:

[0149] Step S4.1. Further, the current recursive network (and existing patents) directly input multiple state characteristics of the battery (such as SOC, SOH, charge and discharge rates, etc.) without screening (the scenario is the same as above, and in fact, complex calculations are difficult to apply in real scenarios), which makes it difficult for the network to fully utilize the impact of these characteristics on the battery state. Different battery characteristics have different contributions under different environmental conditions, and the existing methods fail to dynamically process these characteristics, which easily causes feature redundancy or irrelevant features to affect the performance of the model. Therefore, the battery characteristics are screened based on entropy weights before the recursive network to screen out the features that contribute most to the prediction of battery life and performance. This not only reduces the interference caused by irrelevant features, but also makes the model more focused on the processing of key features, thereby improving the prediction accuracy of the battery state. In this embodiment, step S4.1 specifically includes the following steps:

[0150] Step S4.1.1: First, the input environmental features and monitoring device features have been optimized in step S2. The options in this part are the joint entropy of the overall three-way interaction in step S2 and the features with the strongest interaction with the wide temperature battery information in the environment and monitoring device matrix:

[0151]

[0152] Step S4.1.2: For each wide temperature battery feature X i Carry out entropy weight w i The calculation formula is:

[0153]

[0154] In the formula, p ij is the frequency distribution of the i-th feature at the j-th time step, λ·Var(X i ) Considering the variance of the feature, γ· The dependency between features is measured by partial derivatives, and it is proposed to improve the processing ability of multi-dimensional interactive features.

[0155] The feature set after screening is:

[0156]

[0157] Step S4.2, general recursive networks usually directly process the multi-dimensional features of the input, ignoring the different sources of influence between the features; for example, the environmental features and load features have completely different impact mechanisms on the battery, but the existing technology often mixes these features in the same input, making it difficult for the model to accurately capture the different effects between the features. Therefore, before feature extraction, a dual-branch recursive network architecture is proposed. The dual-branch recursive network architecture can independently process environmental features and monitoring device features, and model these features interactively with battery features respectively. Each branch focuses on one type of feature to ensure that the model can distinguish and refine the different impacts of the environment and load on the battery, and is intended to enhance feature discrimination:

[0158] The environmental feature branch formula is:

[0159]

[0160] Where ⊙ represents the Hadamard product, which is used for element-wise multiplication of environmental features and battery features to capture the local interaction between them;

[0161] The characteristic branch formula of the monitoring device is:

[0162]

[0163] in represents the convolution operation, which is used to capture the local spatiotemporal dependencies between monitoring interaction features and battery features.

[0164] Step S4.3, in the scenario of battery weather resistance monitoring, in fact, according to the actual feedback of battery characteristics, environmental conditions and load change rates are uneven; therefore, if a fixed time step is used like general technology, the model will react slowly when encountering a rapidly changing battery weather resistance monitoring scenario, and will waste computing resources when the characteristics change steadily, which makes it difficult for the model to maintain high efficiency and accuracy in a complex dynamic environment. Therefore, based on this point, a learnable time step operator based on the scenario of battery weather resistance monitoring is proposed. The network dynamically adjusts the time step according to the current input features and hidden states, thereby achieving the goal of speeding up the sampling frequency when the characteristics change drastically, and reducing unnecessary updates when the characteristics change steadily, thereby improving the overall efficiency and accuracy of the prediction:

[0165] Δt t+1 =Δt t ·(1+σ(W time h t +b time ))

[0166] In the formula, Δt t+1 is the step size of the next time step; W time and b time is a learnable parameter used to adjust the step size, h t is the hidden state of the current time step, representing the features captured by the network at the current moment, and σ(·) is the Sigmoid function, which is used to smooth the adjustment of the time step;

[0167] That is, through W time h t +b time The adjustment range of the current time step is calculated, and the model dynamically adjusts the time step according to the change rate of the current feature. If the feature changes greatly, the time step is shortened; otherwise, the time step is extended.

[0168] Step S4.4: In the above-mentioned dual-branch recursive network architecture, the environmental features and monitoring features are processed as independent inputs to perform feature extraction. The present invention proposes a dynamic fusion architecture to ensure that when the environment changes drastically, the weight of the environmental branch is increased, and when the monitoring branch fluctuates greatly, the weight of the monitoring branch is appropriately increased to improve the model's prediction accuracy for battery life and performance:

[0169]

[0170] In the formula, The output of the environment branch focuses on capturing the impact of environmental characteristics on the battery state. To monitor the output of the branch, we focus on the impact of the fluctuation of the monitoring device on the battery status. α(t) and β(t) are dynamic weights that can adjust the contribution of the two branches to the final output in real time according to the current input features.

[0171] Step S4.5: Finally, the output head uses a dual output head for output, and the entropy value prediction formula is:

[0172]

[0173] By integrating the comprehensive characteristics of the environment and load branch output, the system can predict the entropy value of the battery at the next moment to evaluate the stability of the battery state;

[0174] Battery life prediction formula:

[0175]

[0176] The system can predict the remaining life of the battery based on the fused feature output;

[0177] The loss function for entropy value prediction is defined as:

[0178]

[0179] In the formula, is the predicted entropy of the model at time step t, H real (t) is the entropy value actually calculated at time step t, and T is the total number of time steps;

[0180] The loss function for battery life prediction is defined as:

[0181]

[0182] In the formula, is the battery life predicted by the model at time step t, L real (t) is the actual measured battery life at time step t, and T is also the total number of time steps.

[0183] It is further explained that the actual entropy value calculation formula in the prior art is:

[0184] The calculation of the actual entropy value is divided into two main parts: the entropy of the interaction between the environment and the battery and the entropy of the interaction between the monitoring device and the battery. The joint entropy of the strongest feature of the environment interaction and the battery:

[0185] Based on the features in the screened environment matrix that have the strongest interaction with the battery and battery matrix Calculate the joint entropy H of the two env-battery .

[0186]

[0187] Combined entropy of monitoring device and battery:

[0188] Similarly, by monitoring the features in the device matrix that interact most strongly with the battery and battery matrix To calculate the joint entropy H of the two device-battery .

[0189]

[0190] The comprehensive calculation of the actual entropy value is to directly add the above two independent joint entropy values ​​as the time series interaction joint entropy of the entire system:

[0191] H real (t) = α·H env-battery (t)+β·H device-battery (t)

[0192] Step S5: If the entropy value continues to increase, perform a backtracking analysis. Based on the prediction results of step S4, when the prediction results of the system show that the battery life is decreasing or the entropy value is significantly increasing, indicating that the system has indeed entered an unstable state or the weather resistance (temperature resistance) of the battery is decreasing, a backtracking analysis is performed to find potential influencing factors. In this embodiment, step S5 specifically includes the following steps:

[0193] Step S5.1, the system automatically backtracks to check the environmental features that interact most strongly with the battery matrix, especially the temperature, temperature gradient, maximum temperature and minimum temperature. If the ambient temperature is too high or too low, the system backtracks the relationship between the temperature change and the battery health status to identify whether there is a specific temperature that causes the accelerated decay of the battery life;

[0194] Step S5.2, trace back the interaction record between the monitoring device characteristics and the battery to check whether there is a specific load change or power demand that causes a decrease in battery life. Frequent high-load use will cause the battery life to decline rapidly.

[0195] Step S5.3: If the joint entropy value predicted by the system continues to rise, it indicates that the interaction uncertainty between the battery, the environment and the monitoring device increases. At this time, trace back the interaction pattern between the three to check whether the battery is in some extreme environment (such as extremely high temperature or extremely low temperature), resulting in a significant decrease in battery performance.

[0196] Step S5.4, based on the above backtracking results, when the backtracking analysis finds that specific environment or load conditions cause the battery life to decline rapidly, if it is in a high temperature environment, the discharge rate is reduced, if it is in a low temperature environment, the charging efficiency is improved, and the battery preheating is increased; and when it is predicted that the entropy uncertainty continues to increase according to the backtracking results, the monitoring frequency and feature analysis are directly encrypted, the power demand of the transmission line monitoring device is allocated to multiple batteries or different time periods, the impact of a single high load on the battery life is reduced, and the power wall is automatically set.

[0197] The high-weather-resistant wide-temperature battery monitoring method based on the transmission line online monitoring device is applied to the batteries of the transmission line monitoring device in extreme environments, especially in areas with large temperature differences and harsh environments, such as Xinjiang. The high-weather-resistant wide-temperature battery monitoring method based on the transmission line online monitoring device monitors the status data of the battery, environment and monitoring device in real time. On the one hand, it optimizes the performance monitoring and life management of the wide-temperature battery to avoid damage to the battery due to excessive discharge or charging at extreme temperatures; on the other hand, the system accurately evaluates the interactive effects of the environment and monitoring equipment through a double-branch recursive network architecture, significantly improving the accuracy of battery status prediction. In addition, the high-weather-resistant wide-temperature battery monitoring method based on the transmission line online monitoring device also has the ability to adaptively adjust the monitoring frequency to ensure that the battery life is maximized in complex environments. It also achieves efficient information processing through dynamic adjustment, thereby improving the overall stability and efficiency of the transmission line monitoring system.

[0198] Compared with the prior art, the high-weatherability and wide-temperature battery monitoring method based on the online monitoring device for transmission lines has the following advantages: (1) It adopts a multi-dimensional feature screening and entropy weight optimization model. Existing battery monitoring systems usually use matrix splicing to process the multi-dimensional data interaction of batteries, environments and devices, resulting in excessive information redundancy and computational complexity, especially in a wide temperature environment. The high-weatherability and wide-temperature battery monitoring method based on the online monitoring device for transmission lines proposes a feature screening and weight optimization model based on information entropy. By calculating the mutual information value between battery features and environmental and monitoring device features, the most influential feature dimensions are screened out to avoid the high-dimensional data redundancy problem caused by directly splicing all dimensions. Then, the key features screened out are weighted through the entropy weight optimization model to ensure that the most important dimensions in information interaction are given priority. Finally, the model makes entropy calculation more efficient by reducing redundant information, while improving the accuracy of battery status monitoring. (2) It adopts a dynamic time step adjustment mechanism. Most traditional battery monitoring systems use a fixed time step, which cannot flexibly respond to dynamic changes in battery performance and environment, resulting in insufficient monitoring frequency or waste of resources. This high-weather-resistant wide-temperature battery monitoring method based on the transmission line online monitoring device proposes a learnable time step adjustment mechanism. Based on the changes in the current input features and hidden states, the time step is dynamically adjusted through a recursive neural network. When the battery performance or environmental conditions change drastically, the system will shorten the time step and increase the monitoring frequency to capture the abnormal conditions of the battery in a timely manner; when the conditions are stable, the system will extend the time step and reduce the monitoring frequency, thereby improving the monitoring efficiency of the system. This dynamic step adjustment enables the system to predict the battery status more efficiently and accurately in complex environments. (3) Adopting a dual-branch recursive neural network architecture. In the process of battery status prediction, the impact mechanisms of environmental characteristics and monitoring equipment characteristics on the battery are completely different. Existing systems often mix these features and it is difficult to fully capture their independent effects. This high-weather-resistant wide-temperature battery monitoring method based on the transmission line online monitoring device innovatively proposes a dual-branch recursive neural network architecture to process environmental characteristics and monitoring equipment characteristics respectively, and independently model them in combination with battery characteristics. Each branch focuses on one type of feature to ensure that the model can accurately capture the different impacts of environmental conditions and monitoring equipment on the battery status. In addition, the system introduces a dynamic fusion mechanism in the model output stage, which can adjust the weights of the two branches in real time according to the current input characteristics, thereby improving the accuracy of the model's prediction of battery life and performance. (4) Adopting a wide-temperature battery feedback information flow optimization and joint entropy evaluation model. In order to further improve the system's ability to predict battery performance under extreme temperature conditions, this high-weatherability wide-temperature battery monitoring method based on the transmission line online monitoring device introduces a wide-temperature battery feedback information flow optimization model.At each time step, the system dynamically selects the most valuable information flow between the battery, environment, and monitoring equipment to avoid directly splicing all information for entropy calculation, thereby reducing the computational complexity brought by high-dimensional data. Through this optimization, the system can more flexibly adjust the influence of each feature dimension to ensure more efficient multi-dimensional entropy calculation. Finally, through the joint entropy evaluation model, the performance of the battery under wide temperature conditions is globally evaluated to capture abnormal conditions that may occur in the battery in the early stage.

[0199] Embodiment 2: This embodiment provides a high weather resistance and wide temperature range battery monitoring system based on a power transmission line online monitoring device, comprising:

[0200] The monitoring environment initialization module is used to initialize the monitoring environment, connect the wide temperature battery monitoring device, set the monitoring parameters, including the monitoring frequency, the initial state of the battery and the environmental conditions, and ensure the normal operation of the device;

[0201] The entropy feedback information flow optimization module integrates a multi-dimensional entropy feedback mechanism to dynamically select the feature dimensions with the strongest information interaction based on the monitoring data flow and environmental conditions, avoid redundant data, improve monitoring accuracy, and output the optimized monitoring data flow;

[0202] The real-time monitoring execution module is connected to the monitoring environment initialization module and is used to perform full-range monitoring of the wide-temperature battery according to the optimized data stream, and obtain the working parameters of the battery under different ambient temperatures, load conditions and power requirements, where the working parameters include key data such as battery voltage, charge and discharge rate, SOC, internal resistance, etc.;

[0203] The data processing module is used to sort and normalize the data obtained by the real-time monitoring execution module, generate a battery feature matrix, an environmental feature matrix and a monitoring device matrix, and input these data into a multidimensional entropy calculation model for feature analysis;

[0204] A joint entropy calculation module is used to calculate the joint entropy value between the battery, the environment and the monitoring device based on the matrix generated by the data processing module, reflect the stability of the battery system and its complexity under different environmental conditions, and generate a real-time risk assessment report based on the joint entropy value;

[0205] Dynamic warning and feedback module, used to determine whether the battery system is in an unstable state based on the comparison of the joint entropy value and the preset threshold; when the joint entropy value exceeds the dynamic threshold, the warning mechanism is triggered to perform dynamic adjustments to the battery charge and discharge rate or power distribution to prevent the wide temperature battery from premature failure due to extreme conditions;

[0206] The display module is used to display various real-time data, joint entropy analysis results, risk assessment reports and the operating status of the monitoring device during the battery monitoring process, and provide a visual prediction of the battery health trend.

[0207] The high-weather-resistant wide-temperature battery monitoring system based on the online monitoring device of the transmission line is applied to the batteries of the transmission line monitoring device in extreme environments, especially in areas with large temperature differences and harsh environments, such as Xinjiang. The high-weather-resistant wide-temperature battery monitoring system based on the online monitoring device of the transmission line optimizes the performance monitoring and life management of the wide-temperature battery by real-time monitoring of the status data of the battery, environment and monitoring device, and avoids damage to the battery due to excessive discharge or charging at extreme temperatures; on the other hand, the system accurately evaluates the interactive effects of the environment and monitoring equipment through a double-branch recursive network architecture, significantly improving the accuracy of battery status prediction. The high-weather-resistant wide-temperature battery monitoring system based on the online monitoring device of the transmission line also has the ability to adaptively adjust the monitoring frequency to ensure that the battery life is maximized in complex environments. It also realizes efficient information processing through dynamic adjustment, improving the overall stability and efficiency of the transmission line monitoring system.

[0208] Embodiment 3: This embodiment provides a terminal device, which includes a memory, a processor, a communication interface and a communication bus. The memory stores a program that can be run on the processor. When the processor executes the program, the high weather resistance and wide temperature battery monitoring method based on the transmission line online monitoring device in the above embodiment is implemented.

[0209] The processor may be a central processing unit, or other general-purpose processors, digital signal processors, application-specific integrated circuits, field programmable gate arrays or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0210] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs, non-transient computer executable programs and units, such as the corresponding program units in the above method embodiments of the present invention. The processor executes various functional applications and data processing of the processor by running the non-transient software programs, instructions and modules stored in the memory, thereby realizing the high weather resistance and wide temperature battery monitoring method based on the transmission line online monitoring device in the above embodiment.

[0211] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. The memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. One or more programs are stored in the memory, and when executed by the processor, the high weather resistance and wide temperature battery monitoring method based on the transmission line online monitoring device in the above embodiment is executed.

[0212] Embodiment 4: This embodiment provides a storage medium for computer-readable storage, wherein the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement a high weather resistance and wide temperature battery monitoring method based on a power transmission line online monitoring device as in the above embodiment.

[0213] The storage medium may be an internal storage unit of the terminal device, such as a hard disk or memory of the terminal device. The storage medium may also be an external storage device of the terminal device, such as a plug-in hard disk, smart memory card, secure digital card, flash memory card, etc. equipped on the terminal device.

[0214] The above technical features constitute the embodiments of the present invention, which have strong adaptability and implementation effect. Non-essential technical features can be added or reduced according to actual needs to meet the requirements of different situations.

Claims

1. A high weather resistance and wide temperature range battery monitoring method based on an online monitoring device for a transmission line, characterized in that The following steps are involved: Step S1, collecting voltage, current, power and environmental data in real time; Step S2, constructing a time series data matrix of monitoring device, environment and battery characteristics; Step S3, dynamically adjusting the threshold according to the joint entropy and historical status; Step S4, performing refined wide temperature battery feedback analysis; Step S5: If the entropy value continues to increase, perform backtracking analysis.

2. The high weather resistance and wide temperature range battery monitoring method based on the transmission line online monitoring device according to claim 1 is characterized in that In step S1, the power requirement of the monitoring device is set to P device (t), then P device-1 (t)=V device (t)·I device (t) Where V device (t) and I device (t) are the voltage and current requirements of the monitoring device at time t, respectively; Considering the frequency change monitoring device power demand is expressed as: P device (t)=P0+α·f(x) Where P0 is the power consumption during normal operation, α is the proportional constant between frequency and power consumption, and f(x) is the monitoring frequency of the monitoring device; Construct a monitoring device power change rate, which represents the rate of change of the device power demand: Build environmental conditions, including real-time temperature T(t), temperature gradient The maximum temperature of the day T max and the minimum temperature T min ; Among them, the temperature gradient It represents the rate of change of temperature over time, that is: The real-time operating parameters of the wide temperature battery include the battery voltage V b (t), current I b (t), remaining charge SOC(t), battery temperature T b (t), real-time battery output power P b (t), battery charge and discharge rate and the internal resistance of the battery; Among them, the remaining power SOC(t) is calculated according to the rated capacity and current change of the battery: In the formula, C r is the rated capacity of the battery, Δt is the time interval; Among them, the charge and discharge rate Calculated based on the battery current changes: In the formula, C r is the rated capacity of the battery.

3. The high weather resistance and wide temperature range battery monitoring method based on the transmission line online monitoring device according to claim 2 is characterized in that In step S2, the time series data matrix X of the monitoring device is constructed device (t), time series data matrix X of environmental conditions env (t) and the battery characteristic matrix X battery (t): Construct the joint entropy of the wide temperature battery timing joint matrix: Among them, each matrix dimension is the monitoring device matrix X device :n×4, environmental condition matrix X env :n×4, battery feature matrix X battery :n×7; The battery interacts with the environment information and calculates the battery status X battery and the environmental state X env The information entropy between battery ;X env ), find out the most influential dimensions of environmental characteristics: In the formula, represents the i-th feature in the environment matrix; That is, the information entropy of each environmental dimension is calculated by the above formula, and the dimension with the highest information entropy value is selected as the main factor for weighted calculation; Calculate battery status X battery and monitoring device X device The information entropy between battery ;X device ), identify the dimensions in the monitoring device matrix that have the greatest impact on the battery system: In the formula, represents the i-th feature in the monitoring device matrix; Compare the information entropy values ​​of different features in the monitoring device and select the dimension with the greatest impact; Filter out the feature dimension with the largest weight: According to the calculated information entropy value, the feature dimensions with the strongest information interaction, that is, the features with the largest information entropy value, are selected from the environment and monitoring device matrices respectively: Filter the feature dimension with the largest weight from the monitoring device matrix: Based on the calculation of mutual information, the feature dimension with the largest weight in the environment and monitoring device matrix is ​​selected for the calculation of weighted joint entropy; That is, screening after calculating mutual information: sort the mutual information of each dimension and select the feature with the largest mutual information value; For the selected feature dimensions, features with higher information entropy will receive greater weights: in, is the weight of the i-th feature of the environment matrix, normalized based on its information entropy value with the battery; is the weight of the i-th feature of the monitoring device matrix, normalized based on its information entropy value with the battery; Finally, the joint entropy of the temporal interaction of the three is calculated: in, is the weight value of the feature dimension with the strongest information interaction selected from the environment matrix, is the weight value of the feature dimension with the strongest information interaction selected from the monitoring device matrix, is the joint probability distribution between battery features, environmental features, and device features, is the joint entropy.

4. The high weather resistance and wide temperature storage battery monitoring method based on the transmission line online monitoring device according to claim 2 or 3 is characterized in that In step S3, the calculation formula for dynamically adjusting the threshold is: In the formula, H threshold (t) is the dynamic threshold at time t, H baseline is the initial baseline entropy value of the system, γ is the adjustment coefficient, which is used to control the sensitivity of the threshold, N is the length of the time window, which means considering the entropy value history of the past N time steps, H weighted (i) is the weighted joint entropy calculated at time step i; Current weighted joint entropy H weighted (t) exceeds the dynamic threshold H threshold (t), then based on the subsequent step S4, refined wide temperature battery feedback is performed: H weighted (t)≥H threshold (t) 5. The high weather resistance and wide temperature range battery monitoring method based on the transmission line online monitoring device according to claim 2 or 3 is characterized in that Step S4 specifically includes the following steps: Step S4.1, screening the battery features based on entropy weights before the recursive network to select the features that contribute most to the prediction of battery life and performance; Step S4.2, before feature extraction, a dual-branch recursive network architecture is proposed, which can independently process environmental features and monitoring device features, and interactively model these features with battery features respectively, with each branch focusing on one type of feature to ensure that the model can distinguish and refine the different effects of the environment and load on the battery, and is intended to enhance feature discrimination; Step S4.3, dynamically adjust the time step according to the current input features and hidden states through the network, so as to speed up the sampling frequency when the features change drastically, and reduce unnecessary updates when the features change steadily, thereby improving the efficiency and accuracy of the prediction as a whole; Step S4.4, in a dual-branch recursive network architecture, the environmental features and the monitoring features are processed as independent inputs for feature extraction; Step S4.5, the final output head uses a dual output head for output; by integrating the comprehensive characteristics of the environment and load branch output, the system can predict the entropy value of the battery at the next moment to evaluate the stability of the battery state; the system can predict the remaining life of the battery based on the fused feature output.

6. The high weather resistance and wide temperature range battery monitoring method based on the transmission line online monitoring device according to claim 2 or 3 is characterized in that Step S5 specifically includes the following steps: Step S5.1, the system automatically backtracks to check the environmental features that interact most strongly with the battery matrix, especially the temperature, temperature gradient, maximum temperature and minimum temperature. If the ambient temperature is too high or too low, the system backtracks the relationship between the temperature change and the battery health status to identify whether there is a specific temperature that causes the accelerated decay of the battery life; Step S5.2, trace back the interaction record between the monitoring device characteristics and the battery to check whether there is a specific load change or power demand that causes a decrease in battery life. Frequent high-load use will cause the battery life to decline rapidly. Step S5.3: If the joint entropy value predicted by the system continues to rise, it indicates that the interaction uncertainty between the battery, the environment and the monitoring device increases. At this time, the interaction mode between the three is traced back to check whether the battery is in some extreme environment, resulting in a significant decrease in battery performance; Step S5.4, based on the above backtracking results, when the backtracking analysis finds that specific environment or load conditions cause the battery life to decline rapidly, if it is in a high temperature environment, the discharge rate is reduced, if it is in a low temperature environment, the charging efficiency is improved, and the battery preheating is increased; and when it is predicted that the entropy uncertainty continues to increase according to the backtracking results, the monitoring frequency and feature analysis are directly encrypted, the power demand of the transmission line monitoring device is allocated to multiple batteries or different time periods, the impact of a single high load on the battery life is reduced, and the power wall is automatically set.

7. A high weather resistance and wide temperature range battery monitoring system based on an online monitoring device for transmission lines, characterized in that include: The monitoring environment initialization module is used to initialize the monitoring environment, connect the wide-temperature battery monitoring device, set monitoring parameters, and ensure the normal operation of the device; The entropy feedback information flow optimization module is used to dynamically select the feature dimensions with the strongest information interaction based on the monitoring data flow and environmental conditions, avoid redundant data, improve monitoring accuracy, and output the optimized monitoring data flow; The real-time monitoring execution module is used to perform full-range monitoring of the wide-temperature battery according to the optimized data stream, and obtain the working parameters of the battery under different ambient temperatures, load conditions and power requirements; The data processing module is used to sort and normalize the data obtained by the real-time monitoring execution module, generate a battery feature matrix, an environmental feature matrix and a monitoring device matrix, and input these data into a multidimensional entropy calculation model for feature analysis; A joint entropy calculation module is used to calculate the joint entropy value between the battery, the environment and the monitoring device based on the matrix generated by the data processing module, reflect the stability of the battery system and its complexity under different environmental conditions, and generate a real-time risk assessment report based on the joint entropy value; A dynamic early warning and feedback module is used to determine whether the battery system is in an unstable state based on the comparison between the joint entropy value and the preset threshold; The display module is used to display various real-time data, joint entropy analysis results, risk assessment reports and the operating status of the monitoring device during the battery monitoring process, and provide a visual prediction of the battery health trend.

8. A terminal device, comprising a memory and a processor, wherein the memory stores a program that can be run on the processor, characterized in that: When the processor executes the program, the high weather resistance and wide temperature storage battery monitoring method based on the transmission line online monitoring device as described in any one of claims 1 to 6 is implemented.

9. A storage medium, characterized in that: The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the high weather resistance and wide temperature battery monitoring method based on the transmission line online monitoring device as described in any one of claims 1 to 6.

Images