Fault prediction method of battery recovery equipment based on digitization

Abstract

The invention relates to the technical field of equipment fault prediction, and discloses a fault prediction method of battery recovery equipment based on digitization. According to the method, multi-dimensional data of the battery recovery equipment is collected in real time and stored in a cloud end, the running state of the equipment is comprehensively monitored, sufficiency and timeliness of the data are ensured, in the data preprocessing stage, a high-quality preprocessing data set is generated through denoising, missing value filling and standardized processing, and the data recovery efficiency is improved. A data basis is provided for subsequent feature calculation and model establishment, key indexes such as a battery health state feature value, a residual electric quantity feature value and charging and discharging efficiency are calculated, so that the health condition of the equipment is comprehensively evaluated, and a mathematical model established based on a feature set and historical fault data has high reliability. According to the invention, real-time fault risk assessment of the battery recovery equipment can be realized, the prediction result can be output in time, the potential fault risk can be identified, the operation efficiency of the battery recovery equipment is improved, and the loss caused by equipment fault shutdown is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of equipment failure prediction, in particular to a failure prediction method for battery recycling equipment based on digitization. BACKGROUND

[0002] With the global attention to renewable energy and the popularity of electric vehicles, battery recycling equipment plays an increasingly important role in resource recycling. These devices not only effectively recycle waste batteries and reduce environmental pollution, but also extract valuable metals and materials to promote resource recycling. However, the normal operation of battery recycling equipment is crucial to ensure recycling efficiency and safety, so real-time monitoring and prediction of its failure is particularly important.

[0003] Existing battery recycling equipment failure prediction techniques rely heavily on periodic inspections and manual monitoring, often failing to achieve real-time assessment of the equipment's operating status, resulting in equipment failures often being discovered only after they occur, thereby reducing the operating efficiency and recycling effectiveness of battery recycling equipment. SUMMARY

[0004] (I) Technical problems solved

[0005] To address the shortcomings of the prior art, the present application provides a failure prediction method for battery recycling equipment based on digitization, which collects and stores multi-dimensional data of battery recycling equipment in real time, including temperature, humidity, voltage, current and other key parameters, and comprehensively monitors the operating status of the equipment to ensure the sufficiency and timeliness of the data. In the data preprocessing stage, through denoising, missing value filling and standardization processing, a high-quality preprocessed data set is generated to provide a data foundation for subsequent feature calculation and model establishment. The calculated battery health status feature value, remaining capacity feature value, and charging and discharging efficiency are key indicators that can provide strong support for failure prediction, enabling a comprehensive assessment of the health status of the equipment. In addition, the mathematical model established based on the feature set and historical failure data makes the failure prediction more scientific and accurate, enabling real-time failure risk assessment of battery recycling equipment, timely output of prediction results, and rapid identification of potential failure risks by operators, as well as the implementation of appropriate measures. This not only improves the operating efficiency of battery recycling equipment and reduces losses caused by downtime, but also optimizes the battery recycling process and improves the effectiveness of resource recycling, solving the above problems.

[0006] (II) Technical solutions

[0007] To achieve the above-mentioned purposes, the present application provides the following technical solutions: a failure prediction method for battery recycling equipment based on digitization, comprising the following steps: S1. Collect real-time data from the battery recycling equipment, including temperature, humidity, voltage, current, charge / discharge cycle count, equipment operating time, historical fault data, and historical maintenance records, and store them in the cloud. S2. After denoising, filling in missing values, and standardizing the data stored in the cloud, a preprocessed dataset is generated. S3. Calculate the battery's health status characteristics, remaining capacity characteristics, charge / discharge efficiency, temperature change rate, and current change rate from the preprocessed dataset and form a feature set; S4. Establish a mathematical model for fault prediction of battery recycling equipment based on feature set and historical fault data; S5. Collect new data from the battery recycling equipment in real time, input it into the fault prediction mathematical model of the battery recycling equipment, conduct fault risk assessment, and output the prediction results.

[0008] Preferably, the formula for denoising the data in the cloud is as follows: ; In the formula, Indicates a point in time The denoising results Indicates a point in time Original data of battery recycling equipment Indicates the size of the filtering window. Indicates the counting subscript.

[0009] Preferably, the formula for calculating the missing values ​​is as follows: ; In the formula, Indicates a point in time The value to fill in, This represents the previous known value. Indicates the next known value. , Indicates the index of a known value, used to determine the location of missing values ​​and the interpolation process.

[0010] Preferably, the formula for data standardization is as follows: ; In the formula, This represents the standardized data value. This represents the data value of the original battery recycling equipment. This represents the minimum value among all samples in the dataset. This represents the maximum value among all samples in the dataset.

[0011] Preferably, the formula for calculating the health status characteristic value of the battery is as follows: ;In the formula, This represents the battery's health status characteristic value. This indicates the internal resistance value in the new battery state. This indicates the internal resistance value of the battery as measured during actual use.

[0012] Preferably, the formula for calculating the remaining power characteristic value is as follows: ; In the formula, This represents the characteristic value of the remaining battery power. Indicates the initial remaining battery power. Indicates time The charging current, Indicates time The discharge current, This indicates the nominal capacity of the battery. This represents the integration parameter.

[0013] Preferably, the formula for calculating the charge / discharge efficiency is as follows: ; In the formula, Indicates charge / discharge efficiency. This represents the energy that can be effectively utilized during the actual discharge process. This indicates the total energy provided during the charging process.

[0014] Preferably, the formula for calculating the rate of temperature change is as follows: ; In the formula, Indicates the rate of temperature change. This indicates the temperature value measured by the battery recycling equipment at the current moment. This indicates the temperature value at the previous time point.

[0015] Preferably, the formula for calculating the rate of change of current is as follows: ; In the formula, Indicates the rate of change of current. This indicates the current value measured by the battery recycling equipment at the current moment. This indicates the current value at the previous time point.

[0016] Preferably, the fault prediction mathematical model of the battery recycling equipment is as follows: ;

[0017] In the formula, This represents the fault prediction result value of the battery recycling equipment. This represents the model's predicted value when all features are zero. , , ... This represents the degree of influence of each feature variable on the prediction of the outcome. , , ... Various indicators representing battery performance, including , Charge and discharge efficiency, temperature change rate, current change rate, and historical fault data. The error term is a random variable that follows a normal distribution.

[0018] Compared with existing technologies, this invention provides a fault prediction method for digitally based battery recycling equipment, which has the following beneficial effects:

[0019] This invention comprehensively monitors the operating status of battery recycling equipment by collecting and storing multi-dimensional data in real time and in the cloud, including key parameters such as temperature, humidity, voltage, and current. This ensures the sufficiency and timeliness of the data. In the data preprocessing stage, noise reduction, missing value filling, and standardization generate a high-quality preprocessed dataset, providing a data foundation for subsequent feature calculations and model building. The calculated key indicators, such as battery health status characteristics, remaining capacity characteristics, and charge / discharge efficiency, provide strong support for fault prediction, enabling a comprehensive assessment of the equipment's health status. Furthermore, the mathematical model built based on the feature set and historical fault data makes fault prediction more scientific and accurate, enabling real-time fault risk assessment of battery recycling equipment and timely output of prediction results. This helps operators quickly identify potential fault risks and take corresponding measures, which not only improves the operating efficiency of battery recycling equipment and reduces losses caused by downtime due to faults, but also optimizes the battery recycling process and enhances the effectiveness of resource recovery. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the method steps of the present invention. Detailed Implementation

[0021] 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.

[0022] Existing fault prediction technologies for battery recycling equipment largely rely on periodic inspections and manual monitoring, often failing to provide real-time assessment of equipment operating status. This results in faults being detected only after they occur, reducing the operational efficiency and recycling effectiveness of battery recycling equipment. Therefore, a digital-based fault prediction method for battery recycling equipment is proposed. Please refer to [link to relevant documentation]. Figure 1 The method includes the following steps:

[0023] S1. Collect real-time data from the battery recycling equipment, including temperature, humidity, voltage, current, charge / discharge cycle count, equipment operating time, historical fault data, and historical maintenance records, and store them in the cloud.

[0024] During the real-time data collection process of the battery recycling equipment, the operating status of the equipment is first monitored through a high-precision sensor network to obtain key parameters in real time, including temperature, humidity, voltage and current. These sensors adopt industrial-grade standards to ensure the accuracy and stability of the data. At the same time, the number of charge and discharge cycles and the running time of the equipment are also recorded through the built-in counter and clock module. In order to fully understand the health status of the equipment, the system will also periodically access historical fault data and maintenance records stored locally or in the cloud.

[0025] After data collection, the data is uploaded to cloud storage via the secure communication protocol HTTPS using Internet of Things (IoT) technology to ensure the security and integrity of the data during transmission. The cloud data storage uses a distributed database (Amazon DynamoDB) to achieve high availability and scalability. In this way, the real-time data of the battery recycling equipment can not only be collected and stored efficiently, but also provide a solid foundation for subsequent data analysis and fault prediction.

[0026] S2. After denoising, filling in missing values, and standardizing the data stored in the cloud, a preprocessed dataset is generated.

[0027] The formula for denoising data in the cloud is as follows: ; Denoising can significantly reduce random noise in a signal, improve the reliability of the data, and ensure that subsequent analysis and model training are based on accurate data. In the formula, Indicates a point in time The denoising results Indicates a point in time Original data of battery recycling equipment Indicates the size of the filtering window. The count index represents the noise data that may cause overfitting of the machine learning model, thereby reducing the prediction performance, and the denoising can improve the generalization ability of the model, especially in complex environments;

[0028] The calculation formula for filling in missing values is as follows: ;

[0029] In the data collection process, missing values are a common problem, and filling in missing values can avoid the loss of key features due to data missing, thereby retaining more information for analysis and model training, and in the formula, represents the value filled in at time point , represents the previous known value, represents the next known value, , represents the index of the known value, used to determine the position of the missing value and the interpolation process, filling in missing values can prevent the model from being misinterpreted during training, ensuring that the model weights are reasonably updated, thereby improving prediction accuracy;

[0030] The formula for data standardization is as follows: ; Standardization allows data of different dimensions to be compared with each other, which helps to improve the contribution of features to model training, ensuring that the model does not rely solely on the range of feature values of a certain feature, and in the formula, represents the standardized data value, represents the original data value of the battery recycling equipment, represents the minimum value of all samples in the data set, represents the maximum value of all samples in the data set, and the standardized data is more stable in the propagation process, especially when using optimization algorithms such as gradient descent, which can significantly reduce training errors;

[0031] S3, calculate the battery health state feature value, remaining capacity feature value, charge and discharge efficiency, temperature change rate and current change rate from the preprocessed data set and form a feature set; The calculation formula of the battery health state feature value is as follows: ; The health state feature value provides an intuitive indicator of battery performance, which helps to monitor the health of the battery in real time and discover problems in a timely manner, and in the formula, represents the health state feature value of the battery, represents the internal resistance value under the state of a new battery, This indicates the internal resistance value of the battery measured during actual use. By tracking changes in health status characteristics, potential fault risks can be identified in advance, reducing the impact of sudden failures on equipment and its operation. The formula for calculating the characteristic value of remaining battery power is as follows: ; The remaining battery power characteristic value enables equipment operators to accurately determine the current operating status of the battery, ensuring that the equipment operates within an appropriate power range. In the formula, This represents the characteristic value of the remaining battery power. Indicates the initial remaining battery power. Indicates time The charging current, Indicates time The discharge current, This indicates the nominal capacity of the battery. This represents the integral parameter. By properly managing the remaining power, it is possible to avoid the damage to the battery caused by over-discharging or over-charging, thereby extending the battery's lifespan. The formula for calculating charge / discharge efficiency is as follows: ; Charge-discharge efficiency characteristics provide a basis for monitoring the energy conversion efficiency of a battery, helping to assess the battery's current health status. Furthermore, monitoring charge-discharge efficiency can optimize charging strategies and reduce energy waste. (The formula is missing from the original text.) Indicates charge / discharge efficiency. This represents the energy that can be effectively utilized during the actual discharge process. This indicates the total energy provided during the charging process; The formula for calculating the rate of temperature change is as follows: ; By analyzing temperature changes, the optimal operating temperature range is determined, thereby adjusting the equipment's operating environment to ensure the battery operates in its best condition. (The formula is incomplete.) Indicates the rate of temperature change. This indicates the temperature value measured by the battery recycling equipment at the current moment. This indicates the temperature value at the previous time point. Temperature changes have a significant impact on battery performance, and monitoring can help predict performance degradation in advance, allowing for timely intervention. The formula for calculating the rate of change of current is as follows: ; The rate of change of current can reflect the energy consumption trend of equipment during use, and helps to assess load changes and battery adaptability. In the formula, Indicates the rate of change of current. This indicates the current value measured by the battery recycling equipment at the current moment. represents the current value at the previous time point, sudden current changes are often early indicators of faults, by monitoring this parameter, users can be warned of potential faults in advance; S4, a fault prediction mathematical model of the battery recycling equipment is established according to the feature set and historical fault data; The mathematical model is as follows: ; By establishing a fault prediction model, early warning of equipment failure can be achieved, maintenance or replacement can be performed in advance, downtime can be reduced, and as new data is continuously input and analyzed, the fault prediction model can be continuously updated and optimized to adapt to complex and changing operating environments, achieve self-improvement, and the formula, represents the fault prediction result value of the battery recycling equipment, represents the prediction value of the model when all features are zero, 、 、 、...、 represents the influence degree value of each feature variable on the result prediction, 、 、 、...、 represents each index of battery performance, including 、 , charge and discharge efficiency, temperature change rate, current change rate and historical fault data, represents the error term, which is a random variable subject to normal distribution, this mathematical model is based on a large amount of historical data and feature analysis, improves the scientificity and accuracy of decision-making, and makes equipment management more efficient; S5, real-time collection of new data of the battery recycling equipment, input into the fault prediction mathematical model of the battery recycling equipment, perform fault risk assessment, and output the prediction result; In this process, real-time collection of new data of the battery recycling equipment is achieved through an efficient Internet of Things (IoT) architecture, embedded sensors and data acquisition modules are used to continuously monitor key operating parameters of the equipment, such as voltage, current, temperature and charge and discharge cycle times, etc., these real-time data are quickly uploaded to the cloud through secure communication protocols (such as MQTT or HTTPS), ensuring the safety and integrity of the data during transmission; Once the new data is collected to the cloud, the system will automatically input it into the pre-established failure prediction mathematical model, which is based on historical data, real-time characteristic values and current equipment status, uses a deep learning model to accurately analyze the health status and potential failure risk of the equipment, and through comparison and analysis of new data with historical data, the model can assess the failure risk under the running condition of the equipment in real time and output detailed prediction results, usually including the possibility of failure occurrence, potential failure type and recommended maintenance measures.

[0032] Through the comprehensive application of the above-mentioned method, the failure problems of the battery recycling equipment are found in time, and maintenance measures are taken, which not only improves the operation efficiency of the battery recycling equipment and reduces the loss caused by failure downtime, but also optimizes the battery recycling process and improves the effect of resource recycling.

[0033] Although embodiments of the present application have been shown and described, it is to be understood that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for failure prediction of a digitization-based battery recycling apparatus, characterized by, The method comprises the following steps: S1, collecting real-time data of the battery recycling equipment, including temperature, humidity, voltage, current, charge and discharge cycle number, equipment running time, historical fault data and historical maintenance record data and storing in the cloud; S2, generating a preprocessed data set after denoising, filling missing values and data standardization processing of the data stored in the cloud; S3, calculating the health state characteristic value, residual capacity characteristic value, charge and discharge efficiency, temperature change rate and current change rate of the battery from the preprocessed data set and forming a feature set; S4, establishing a fault prediction mathematical model of the battery recycling equipment according to the feature set and the historical fault data; S5, collecting new data of the battery recycling equipment in real time, inputting the fault prediction mathematical model of the battery recycling equipment, performing fault risk assessment, and outputting the prediction result.

2. The method of claim 1, wherein: The formula for denoising the data in the cloud is as follows: In the formula, denotes the denoising result at time point , denotes the battery recycling equipment original data at time point , denotes the filter window size, denotes the count index.

3. The method of claim 2, wherein: The calculation formula for filling missing values is as follows: In the formula, denotes the value filled in at the time point , denotes the previous known value, denotes the next known value, , denotes the index of the known value, used to determine the position of the missing value and the interpolation process.

4. The method of claim 3, wherein: The formula for data standardization is as follows: In the formula, represents the standardized data value, represents the data value of the original battery recycling equipment, represents the minimum value of all samples in the data set, represents the maximum value of all samples in the data set.

5. The method of claim 4, wherein: The formula for calculating the health state characteristic value of the battery is as follows: In the formula, represents a health state characteristic value of the battery, represents an internal resistance value in a new battery state, represents an internal resistance value measured in actual use of the battery.

6. The method of claim 5, wherein: The formula for calculating the residual capacity characteristic value is as follows: In the formula, This represents the characteristic value of the remaining battery power. Indicates the initial remaining battery power. Indicates time The charging current, Indicates time The discharge current, This indicates the nominal capacity of the battery. This represents the integration parameter.

7. The method of claim 6, wherein: The formula for calculating the charge and discharge efficiency is as follows: In the formula, represents the charging and discharging efficiency, represents the energy that can be effectively utilized in the actual discharging process, represents the total energy provided in the charging process.

8. The method of claim 7, wherein: The formula for calculating the temperature change rate is as follows: In the formula, represents the temperature change rate, represents the temperature value measured by the battery recycling equipment at the current time, represents the temperature value at the previous time point.

9. The method of claim 8, wherein: The formula for calculating the current change rate is as follows: In the formula, represents the current rate of change, represents the current value measured by the battery recycling device at the current time, represents the current value at the previous time point.

10. The method of claim 9, wherein: The fault prediction mathematical model of the battery recycling equipment is as follows: In the formula, represents the failure prediction result value of the battery recycling equipment, represents the prediction value of the model when all features are zero, represents the impact degree value of each feature variable on the result prediction, represents each index of battery performance, including charge and discharge efficiency, temperature change rate, current change rate, and historical failure data, represents the error term, which is a random variable subject to normal distribution.​​​​​​​

Images