Solar street lamp battery residual life detection method

Abstract

The application provides a kind of solar street lamp battery residual life detection method, belongs to solar street lamp technical field, this method is through to solar street lamp battery comprehensive parameter collection and analysis, first obtains battery basis and operating parameter, and carries out discharge capacity test and internal resistance measurement, simultaneously collects battery operating history data.Based on these data, construct battery feature vector, use neural network to establish life prediction model, through training to sample battery.In practical application, through the real-time parameter of the battery to be tested, such as voltage, current, surface temperature and cumulative running time, input the pre-training model with these characteristics, the residual life of the battery can be accurately predicted, which provides a scientific and accurate decision basis for the maintenance and replacement of solar street lamp battery, effectively improves the intelligent level of battery management and system reliability.Solves the technical problem that the reliability is not high according to the service time to judge the residual life of solar street lamp battery.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of solar street lamps, and particularly relates to a solar street lamp battery residual life detection method. BACKGROUND

[0002] As a new type of lighting equipment using solar power generation, the solar street lamp is popular in the fields of urban road lighting and rural area lighting due to its environmental protection and energy saving characteristics. Compared with the traditional street lamp, the power supply mode of the solar street lamp is changed from power grid power supply to independent solar cell power supply, thereby greatly reducing the operation and maintenance cost. However, the solar cell, which is the core component of the solar street lamp, is a key factor affecting the performance and working reliability of the whole system.

[0003] At present, the service life of the solar street lamp battery is generally short, and the average service life is only 5-8 years, which seriously restricts the large-scale popularization and application of the solar street lamp. The main reasons for the short service life of the battery include battery capacity attenuation, internal resistance increase, and cycle life depletion. If the service life of the battery is too short, the replacement frequency of the solar street lamp will be increased, thereby increasing the operation and maintenance cost and weakening the economic advantage of the solar street lamp. Therefore, how to effectively prolong the service life of the solar street lamp battery has become a key technical problem to be solved at present.

[0004] In view of the above problems, the industry generally simply judges the residual life of the solar street lamp battery according to the use time, and the reliability is not high. SUMMARY

[0005] Therefore, the application provides a solar street lamp battery residual life detection method, which can solve the technical problem that the residual life of the solar street lamp battery is often simply judged according to the use time in the prior art, and the reliability is not high.

[0006] The application is implemented as follows:

[0007] The application provides a solar street lamp battery residual life detection method, which comprises the following steps:

[0008] S01, acquiring the basic parameters of a sample solar street lamp battery, including a battery rated capacity value, a battery working voltage value, a battery design life value, and a battery internal structure parameter value;

[0009] S02, collecting real-time operation parameters of the sample solar street lamp battery, including a battery real-time voltage value, a battery real-time current value, a battery environment temperature value, a battery surface temperature value, a battery charging time value, and a battery discharging time value;

[0010] S03, performing a battery discharge capacity test to measure the actual discharge capacity value of the sample solar street lamp battery under standard working conditions.

[0011] S04、acquiring the ratio of the real-time voltage value of the battery and the real-time current value of the battery in a preset sampling period to obtain a battery internal resistance measurement sequence;

[0012] S05, compensating the battery internal resistance measurement sequence based on a temperature compensation equation set to obtain a battery internal resistance compensation value at a standard temperature;

[0013] S06, obtaining the operation history data of the sample solar street lamp battery, including the battery charge-discharge cycle number value, the battery cumulative charge value, the battery cumulative discharge value, the battery historical maximum temperature value, and the battery historical minimum temperature value;

[0014] S07, calculating the battery coulomb efficiency value according to the battery cumulative charge value and the battery cumulative discharge value;

[0015] S08, constructing a battery feature vector by using the battery internal resistance compensation value, the battery coulomb efficiency value, the actual discharge capacity value, and the battery charge-discharge cycle number value;

[0016] S09, establishing a battery life neural network prediction model, inputting the battery feature vector and the corresponding actual remaining life value into the battery life neural network prediction model for training to obtain a trained battery life prediction model;

[0017] S10, collecting the rapid detection parameters of the to-be-tested solar street lamp battery, including the real-time voltage value of the to-be-tested battery, the real-time current value of the to-be-tested battery, and the surface temperature value of the to-be-tested battery;

[0018] S11, calculating the ratio of the real-time voltage value of the to-be-tested battery and the real-time current value of the to-be-tested battery to obtain a to-be-tested battery internal resistance estimation value;

[0019] S12, obtaining the cumulative operation duration value of the to-be-tested solar street lamp battery;

[0020] S13, inputting the to-be-tested battery internal resistance estimation value, the to-be-tested battery surface temperature value, and the cumulative operation duration value into the trained battery life prediction model to obtain a to-be-tested battery remaining life prediction value.

[0021] On the basis of the above technical solution, the solar street lamp battery remaining life detection method of the present application can be further improved as follows:

[0022] The temperature compensation equation set includes a heat conduction equation, a temperature distribution equation, an internal resistance compensation equation, a heat balance equation, and a time response equation.

[0023] Further, the heat conduction equation is used to calculate the temperature conduction relationship between the battery interior and the surface, and the input includes the battery surface temperature value, the battery environment temperature value, the battery internal structure parameter value, and the battery real-time current value, and the output is the battery internal temperature gradient value.

[0024] Further, the temperature distribution equation is used to describe the temperature distribution characteristics of each layer structure of the battery, and the input includes the battery internal temperature gradient value, the battery internal structure parameter value, the battery real-time current value, the battery charging duration value, and the battery discharging duration value, and the output is the battery temperature distribution data.

[0025] Further, the internal resistance compensation equation is used to establish the mapping relationship between the internal resistance and the temperature, and the input includes the battery temperature distribution data, the battery internal resistance measurement sequence, the battery historical maximum temperature value, and the battery historical minimum temperature value, and the output is the battery internal resistance temperature compensation coefficient value.

[0026] Further, the heat balance equation is used to calculate the battery steady-state working temperature, and the input includes the battery environment temperature value, the battery real-time current value, the battery real-time voltage value, and the battery surface temperature value, and the output is the battery balance temperature value.

[0027] Further, the time response equation is used to describe the dynamic characteristics of temperature change, and the input includes the battery balance temperature value, the preset sampling period, the battery real-time current value, and the battery real-time voltage value, and the output is the battery dynamic temperature response data.

[0028] Further, the sample solar street lamp includes at least 100 or more solar street lamps.

[0029] The rapid detection parameters of the to-be-tested solar street lamp battery are collected by a signal collection device, and the signal collection device includes:

[0030] A signal collection unit, the signal collection unit includes:

[0031] A rapid voltage collection sensor, the rapid voltage collection sensor is connected in parallel with the to-be-tested solar street lamp battery;

[0032] A rapid current collection sensor, the rapid current collection sensor is connected in series in the power supply loop of the to-be-tested solar street lamp battery;

[0033] A rapid temperature collection sensor, the rapid temperature collection sensor is fixed on the surface of the to-be-tested solar street lamp battery;

[0034] A signal conditioning unit, the signal conditioning unit includes:

[0035] An operational amplifier circuit is configured to amplify the analog signal output by the second signal acquisition unit;

[0036] A band-pass filter circuit is configured to filter the amplified signal;

[0037] A high-speed analog-to-digital conversion circuit is configured to convert the processed analog signal into a digital signal;

[0038] A control processing unit comprises:

[0039] A low-power single-chip microcomputer circuit is configured to receive the digital signal output by the second signal conditioning unit;

[0040] A fast storage circuit is connected to the low-power single-chip microcomputer circuit;

[0041] A wireless communication circuit is connected to the low-power single-chip microcomputer circuit.

[0042] Further, the battery life neural network prediction model adopts an improved bidirectional long short-term memory network structure, specifically, an attention allocator is added between the memory cell module and the forgetting gate module in the bidirectional long short-term memory network structure, the attention allocator is used to assign different weight coefficients to different feature parameters in the input battery feature vector, thereby highlighting the key parameters that have a greater impact on the remaining life prediction result; the specific structure of the attention allocator is composed of a feedforward neural network, which includes three hidden layers, the first hidden layer is provided with 32 neurons, the second hidden layer is provided with 16 neurons, and the third hidden layer is provided with 8 neurons, the attention allocator calculates the attention weight value of each feature parameter in the input battery feature vector, and multiplies the calculated attention weight value by the original feature parameter to generate a weighted feature vector.

[0043] Further, the improved bidirectional long short-term memory network structure is described in detail as follows: the input layer receives the battery feature vector, which is input to the bidirectional long short-term memory layer after weighted processing by the attention allocator, the bidirectional long short-term memory layer includes a forward-propagating long short-term memory cell and a backward-propagating long short-term memory cell, each long short-term memory cell is provided with 64 neurons, the outputs of the forward-propagating long short-term memory cell and the backward-propagating long short-term memory cell are fused through a fully connected layer, the fully connected layer is provided with 128 neurons, batch normalization processing is adopted to reduce the deviation of data distribution, and finally the battery remaining life prediction value is obtained through an output layer, the output layer adopts a linear activation function.

[0044] Further, the relevant equations or calculations are described in detail as follows:

[0045] 1. The calculation of the battery resistance measurement sequence is specifically represented as follows:

[0046] ;

[0047] In the formula, Rt is the resistance measurement value at time t, with units of ohms; Vt is the real-time voltage value of the battery at time t, with units of volts; It is the real-time current value of the battery at time t, with units of amperes.

[0048] 2. The heat conduction equation is specifically represented as follows:

[0049] ;

[0050] In the formula, T is the temperature field function, with units of kelvin; t is time, with units of seconds; a is the thermal diffusivity, with units of square meters per second; x is the spatial coordinate; Q is the heat source term, which is the Joule heat dissipation generated by the current, with units of watts per cubic meter; c is the specific heat capacity, with units of joules per kilogram kelvin; p is the density, with units of kilograms per cubic meter.

[0051] 3. The temperature distribution equation is specifically represented as follows:

[0052] ;

[0053] In the formula, T (r, t) is the temperature at time t at a distance r from the center of the battery, with units of kelvin; T0 is the surface temperature of the battery, with units of kelvin; Q is the heat source intensity, with units of watts; k is the thermal conductivity, with units of watts per meter kelvin; erfc is the error function; r is the radial distance, with units of meters.

[0054] 4. The internal resistance compensation equation is specifically represented as follows:

[0055] ;

[0056] In the formula, Rt is the temperature-compensated resistance value, with units of ohms; R is the measured resistance value, with units of ohms; a is the temperature coefficient; T is the current temperature in Kelvin; Tref is the reference temperature in Kelvin; Soc is the state of charge correction function; N is the cycle number correction function.

[0057] 5. The heat balance equation is specifically represented as follows:

[0058] ;

[0059] where, m is the battery mass in kilograms; Cp is the specific heat capacity in Joules per kilogram Kelvin; I is the current in Amperes; R is the internal resistance in Ohms; h is the convective heat transfer coefficient in Watts per square meter Kelvin; A is the surface area in square meters; Tambient is the ambient temperature in Kelvin; T is the battery temperature in Kelvin; S is the Stefan-Boltzmann constant; e is the emissivity.

[0060] 6. The battery Coulombic efficiency value is specifically represented as follows:

[0061] ;

[0062] where, C is the Coulombic efficiency in percent; Idischarge is the discharge current in Amperes; Icharge is the charge current in Amperes; tstart is the discharge start time in seconds; tend is the charge start time in seconds.

[0063] 7. The time response equation is specifically represented as follows:

[0064] ;

[0065] where, T is the battery temperature at time t in Kelvin; T is the system time constant in seconds; K is the gain coefficient; Tambient is the ambient temperature in Kelvin; P is the input power in Watts; m is the battery mass in kilograms; Cp is the specific heat capacity in Joules per kilogram Kelvin.

[0066] 8. The battery feature vector is constructed as follows:

[0067] ;

[0068] wherein, R is the compensated internal resistance value in ohm; η is the coulomb efficiency in percentage; C is the actual discharge capacity value in ampere-hour; N is the number of charge-discharge cycles; S is the state of health value in percentage; Tmax is the historical maximum temperature in Kelvin; Tmin is the historical minimum temperature in Kelvin; D is the depth of discharge in percentage.

[0069] 9. The attention weight calculation is constructed as follows:

[0070] ;

[0071] ;

[0072] wherein, wi is the attention weight of the i-th feature; ei is the energy score of the i-th feature; a is the attention vector; W is the weight matrix; x is the input feature; b is the bias term; N is the number of features.

[0073] 10. The forward propagation calculation of the improved bidirectional long short-term memory network is constructed as follows:

[0074] ;

[0075] ;

[0076] ;

[0077] ;

[0078] ;

[0079] ;

[0080] wherein, f is the forget gate output; i is the input gate output; for the candidate memory cell; for the current memory cell; for the output gate output; for the hidden state; for the weight matrix; for the bias term; for the current input; for the sigmoid activation function.

[0081] The principles and significance of constructing these equations are explained in detail as follows:

[0082] 1. Battery internal resistance measurement equation:

[0083] This equation is based on Ohm's law and uses the ratio of voltage to current to calculate internal resistance. Considering the dynamic characteristics of the battery internal resistance, the instantaneous internal resistance value is obtained by real-time sampling. This method is simple, effective and easy to implement. The ratio of voltage to current reflects the internal impedance characteristics of the battery. The larger the internal resistance, the poorer the performance of the battery. By continuous sampling, the trend of internal resistance change over time can be obtained.

[0084] 2. Heat conduction equation:

[0085] This equation is based on Fourier's heat conduction law and uses a three-dimensional heat diffusion model to describe the temperature field distribution inside the battery. The heat source term is introduced to consider the influence of current-induced Joule heat dissipation. The partial differential term in the equation describes the propagation characteristics of temperature in space, and the heat source term describes the influence of internal heat on the temperature field. This equation can accurately describe the temperature distribution law inside the battery.

[0086] 3. Temperature distribution equation:

[0087] This equation is based on the point source instantaneous solution and error function, and considers the cylindrical structure characteristics of the battery. The temperature distribution is described by the radial distance. The introduction of the error function enables the equation to describe the transient conduction process of temperature. The thermal conductivity and thermal diffusivity in the equation reflect the thermal conductivity characteristics of the material. This equation form is easy to solve and has clear physical meaning.

[0088] 4. Internal resistance compensation equation:

[0089] This equation considers the comprehensive influence of temperature, state of charge and cycle number on internal resistance. A quadratic polynomial is used to describe the temperature influence, and a correction function is introduced to consider other factors. The temperature term uses a quadratic form because the battery internal resistance has a nonlinear relationship with temperature. The introduction of the correction function enhances the adaptability of the equation, making the internal resistance compensation more accurate.

[0090] 5. Heat balance equation:

[0091] The equation is based on the principle of energy conservation, considering three heat transfer modes: focal dissipation, convection heat transfer and radiation heat transfer. The left side of the equation describes the rate of change of battery temperature, and the right side of the equation corresponds to the three heat transfer modes. The focal dissipation term is proportional to the square of the current, which reflects the significant impact of large current on temperature rise. The convection term and the radiation term describe the heat exchange process with the environment.

[0092] 6. Coulomb efficiency calculation equation:

[0093] The equation calculates the coulomb efficiency by the time integral ratio of charge and discharge current. The integral form can accurately reflect the energy conversion efficiency of the entire charge and discharge process. The efficiency value directly reflects the energy conversion performance of the battery, which is an important indicator for evaluating battery performance.

[0094] 7. Time response equation:

[0095] The equation uses a first-order differential equation to describe the dynamic response characteristics of the battery temperature, considering the influence of system inertia and external excitation. The time constant reflects the response speed of the system, the gain coefficient reflects the amplification of the system, and the power term describes the influence of internal heating.

[0096] 8. Battery feature vector construction:

[0097] The vector design uses multi-dimensional features to represent the battery state, including electrochemical characteristics (internal resistance, coulomb efficiency), capacity characteristics (actual discharge capacity), usage history (cycle number), state of health (SOH), temperature characteristics (maximum and minimum temperature), and usage intensity (discharge depth). The selection of each feature is based on its impact on battery life, forming a comprehensive feature space that facilitates the training and prediction of subsequent neural networks.

[0098] 9. Attention weight calculation equation:

[0099] The equation is based on the attention mechanism theory, and the feature weight is calculated by the energy score, which realizes the adaptive adjustment of the importance of different features. The softmax function in exponential form ensures that the weight sum is 1 and all positive values, and the tanh activation function introduces non-linear characteristics to enhance the expression ability of the model. The whole mechanism can highlight the role of key features.

[0100] 10. Improved bidirectional long short-term memory network calculation equation set:

[0101] The equation set describes the core calculation process of LSTM, including the calculation of forget gate, input gate, memory cell update and output gate.

[0102] The forget gate equation determines the retention degree of historical information, and the sigmoid function limits the output to 0-1.

[0103] The input gate equation controls the receiving degree of the current input information, and determines the new information together with the candidate memory unit;

[0104] The memory unit update equation combines the historical information and the new information by weighting, and realizes the modeling of long-term dependence;

[0105] The output gate equation controls the output degree of information, and transmits the processed information to the next moment.

[0106] Compared with the prior art, the solar street lamp battery residual life detection method provided by the present application has the beneficial effects that:

[0107] 1. Quickly and non-destructively obtain key parameters of the battery. The present application uses fast voltage, current and temperature acquisition sensors to collect key state parameters such as the internal resistance and temperature of the battery to be measured in real time and high precision, without disassembling the battery, thereby greatly improving the data acquisition efficiency. This provides a reliable input data basis for subsequent residual life prediction.

[0108] 2. Establish an accurate battery life prediction model. The present application combines battery internal resistance measurement, temperature compensation, machine learning and other core technologies to construct an improved bidirectional long short-term memory neural network model, which can more accurately predict the residual service life of the battery. The model automatically learns the importance of each battery feature parameter to the life prediction through the attention mechanism, greatly improving the accuracy of the prediction results.

[0109] 3. Prolong the service life of the battery and reduce the operation and maintenance cost. Based on the battery residual life prediction results obtained by the method of the present application, the operation and maintenance personnel can timely find the signs of battery aging and reasonably arrange the battery replacement plan in advance, thereby avoiding premature failure of the battery, fully utilizing the service life of the battery and reducing the overall maintenance cost of the solar street lamp system.

[0110] In summary, the present application solves the technical problem that the prior art often simply judges the residual life of the solar street lamp battery according to the use time, and the reliability is not high. BRIEF DESCRIPTION OF DRAWINGS

[0111] Figure 1 The flowchart of the method provided by the present application is shown in the figure;

[0112] Figure 2 The schematic diagram of the solar street lamp battery signal acquisition device in the embodiment of the present application is shown in the figure;

[0113] Figure 3 The curve graph of the change of the internal resistance of the battery with the number of charge and discharge cycles in the embodiment is shown in the figure;

[0114] Figure 4 The performance evaluation result schematic diagram of the battery residual life prediction model is shown in the figure. DETAILED DESCRIPTION

[0115] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application.

[0116] As Figure 1 shown in FIG. 1, which is a flow chart of a method for detecting the remaining life of a solar street lamp battery provided by the present application, the specific embodiments of each step of the present application will be described in detail as follows.

[0117] The specific embodiment of step S01 is to first obtain the battery rated capacity value from the product specification provided by the battery manufacturer, which is usually 20 ampere-hours to 100 ampere-hours, the battery operating voltage value is generally 12 volts to 48 volts, the battery design life value is usually 3 years to 5 years, and the battery internal structure parameter values include the positive electrode material thickness of 0.1 mm to 0.3 mm, the negative electrode material thickness of 0.08 mm to 0.25 mm, the separator thickness of 0.02 mm to 0.05 mm, and the electrolyte concentration of 1 mol / L to 2 mol / L. These basic parameters are verified and confirmed by a special battery parameter tester, the measurement accuracy of the tester should be no less than 0.1%, and the test environment temperature should be maintained at about 25°C, with a relative humidity controlled between 45% and 65%.

[0118] The specific embodiment of step S02 is to monitor the battery operating state in real time through a high-precision data acquisition system, with a sampling frequency set to 1 Hz, the battery real-time voltage value measured by a Hall voltage sensor with a measurement range of 0 V to 60 V and a measurement accuracy of 0.1%, the battery real-time current value measured by a Hall current sensor with a measurement range of 0 A to 50 A and a measurement accuracy of 0.1%, the battery environment temperature value measured by a PT100 temperature sensor with a measurement range of -40°C to 85°C and a measurement accuracy of 0.1°C, the battery surface temperature value measured by an infrared temperature sensor with a measurement range of -40°C to 85°C and a measurement accuracy of 0.1°C, and the battery charging duration value and discharging duration value obtained by a system clock, with a timing accuracy of 1 second. After all the collected data are processed by a signal conditioning circuit, they are converted into digital signals by an analog-to-digital converter with a 16-bit analog-to-digital conversion accuracy and a conversion rate no less than 100 kHz.

[0119] The specific implementation of step S03 is to perform a discharge capacity test using a professional battery test system under standard conditions of an ambient temperature of 25 degrees Celsius and a relative humidity of 60%, first charge the battery to a full state of charge with a charging current of 0.2 times the rated capacity value and a charging cutoff voltage of 2.4 volts per single battery, then stand for 1 hour to make the battery temperature tend to be stable, then discharge at a constant current of 0.2 times the rated capacity value, and the discharge cutoff voltage is 1.8 volts per single battery, and the integral value of the current during the discharge process is recorded as the actual discharge capacity value. During the test, the battery voltage, current and temperature data are recorded every 10 seconds, and the average value of the whole test process repeated for 3 times is taken as the final result.

[0120] The specific implementation of step S04 is to collect the real-time voltage value and the real-time current value of the battery every 1 second within a preset sampling period, and the sampling time is not less than 1 hour. The battery internal resistance measurement value is obtained by dividing the real-time voltage value by the real-time current value, a battery internal resistance measurement sequence containing 3600 data points is formed, a sliding average filtering algorithm is used to smooth the measurement sequence, the sliding window length is set to 10 data points to eliminate the influence of measurement noise, and a median filtering algorithm is used to remove abnormal values, and the filtering window length is set to 5 data points. Finally, the least square method is used to fit the processed data to obtain the battery internal resistance change trend curve.

[0121] The specific implementation of step S05 is to construct a temperature compensation model based on the heat conduction equation, the temperature distribution equation, the internal resistance compensation equation, the heat balance equation and the time response equation. First, the temperature conduction relationship between the internal layers of the battery is calculated by the heat conduction equation, the thermal diffusivity value is in the range of 1.0*10-6 square meters per second to 5.0*10-6 square meters per second, then the temperature distribution equation is used to analyze the radial temperature distribution characteristics of the battery, the thermal conductivity value is in the range of 0.5 watts per meter kelvin to 2.0 watts per meter kelvin, then the internal resistance compensation equation is used to establish the quantitative relationship between the internal resistance and the temperature, the temperature coefficient k1 is in the range of 0.001 to 0.005, and k2 is in the range of 0.0001 to 0.0005, then the heat balance equation is used to calculate the steady-state working temperature of the battery, the convective heat transfer coefficient is in the range of 5 watts per square meter kelvin to 20 watts per square meter kelvin, and finally the time response equation is used to describe the temperature dynamic change characteristics, and the system time constant is in the range of 100 seconds to 300 seconds.

[0122] The specific implementation of step S06 is to extract the battery operation history data from the historical database of the battery management system, the battery charge-discharge cycle number value is obtained by detecting the voltage inflection point in the charge-discharge process, and an effective cycle is recorded as 1 when the discharge depth is greater than 20% and the charge is recovered to more than 90%, the battery cumulative charge value is calculated by time integration of the charging current, the integration time interval is 1 second, the battery cumulative discharge value is calculated by time integration of the discharge current, the integration time interval is 1 second, the battery historical maximum temperature value and the minimum temperature value are obtained by screening from the temperature monitoring data, and the data query time span is the whole operation period since the battery is put into use.

[0123] The specific implementation of step S07 is to calculate the battery coulomb efficiency value according to the battery cumulative charge value and the battery cumulative discharge value, first, the charging current and the discharging current are time-integrated to obtain the total charge-discharge electric quantity, and the trapezoidal integration method is used to improve the calculation accuracy, then the percentage of the total discharge electric quantity to the total charge electric quantity is calculated to obtain the coulomb efficiency value, which is usually between 85% and 95%, if the coulomb efficiency value is lower than 80%, it indicates that the battery performance has obviously declined and needs to be paid attention to, and the influence of current measurement error needs to be considered in the calculation process, generally the relative error of current measurement is required to be not more than 0.5%.

[0124] The specific implementation of step S08 is to construct a battery feature vector after standardizing the battery internal resistance compensation value, the battery coulomb efficiency value, the actual discharge capacity value, the battery charge-discharge cycle number value and other characteristic parameters, the standardization adopts the minimum-maximum value normalization method, and each characteristic parameter is mapped to the interval of 0 to 1, wherein the standardization interval of the internal resistance compensation value is 0.5 milliohm to 5 milliohm, the standardization interval of the coulomb efficiency value is 80% to 100%, the standardization interval of the actual discharge capacity value is 60% to 100% of the rated capacity, and the standardization interval of the charge-discharge cycle number value is 0 to 2000 times, the constructed feature vector has a dimension of 8, and contains main feature information reflecting the battery performance decline.

[0125] The specific implementation of step S09 is to construct a battery life prediction model by using an improved bidirectional long short-term memory network, first, a network structure with attention mechanism is designed, the input layer receives an 8-dimensional battery feature vector, the attention allocator adopts a three-layer feedforward neural network structure, the first layer has 32 neurons, the second layer has 16 neurons, and the third layer has 8 neurons, and the activation function adopts the hyperbolic tangent function, then a bidirectional long short-term memory layer is constructed, which contains 64 memory units in the forward and reverse directions, the network is trained by using the stochastic gradient descent algorithm, the learning rate is set to 0.001, the batch size is set to 32, and the training rounds are set to 500 rounds, finally, the bidirectional outputs are fused through a fully connected layer, 128 neurons are set, and the batch normalization method is used to reduce data distribution deviation.

[0126] The specific implementation of step S10 is to use a signal acquisition device to perform rapid parameter detection on the solar street lamp battery to be measured. The rapid voltage acquisition sensor is implemented by using a high-precision voltage division circuit, the range is 0 volt to 60 volts, the sampling rate is 1 kilohertz, the rapid current acquisition sensor is implemented by using a Hall sensor, the range is 0 ampere to 50 amperes, the sampling rate is 1 kilohertz, the rapid temperature acquisition sensor is implemented by using a thermocouple, the range is minus 40 degrees Celsius to plus 85 degrees Celsius, the sampling rate is 10 hertz, and the output signals of all sensors are processed through a signal conditioning circuit, including operational amplification, band-pass filtering, analog-to-digital conversion and the like.

[0127] The specific implementation of step S11 is to obtain an estimated value of the internal resistance of the battery to be measured by calculating the ratio of the real-time voltage value of the battery to be measured to the real-time current value of the battery to be measured, the acquisition time is not less than 10 minutes, the sampling interval is 1 second, not less than 600 internal resistance estimation data points are obtained, the estimation data is processed by using a Kalman filtering algorithm to remove the influence of measurement noise and abnormal values, the process noise covariance matrix and the measurement noise covariance matrix of the Kalman filtering are obtained through experiments, and finally the mean value of the processed data is calculated as the final internal resistance estimation value.

[0128] The specific implementation of step S12 is to obtain the cumulative running time value of the solar street lamp battery to be measured by querying the running record of the battery management system. The cumulative running time is calculated from the first installation of the battery, and is accurate to hours. The running record includes the working state information of the battery, such as the charging state, the discharging state, the idle state and the like. By analyzing these state information, the time of the non-working state of the battery can be eliminated, and the true cumulative running time value is obtained.

[0129] The specific implementation of step S13 is to input the estimated value of the internal resistance of the battery to be measured, the surface temperature value of the battery to be measured and the cumulative running time value into the trained battery life prediction model. First, the input parameters are standardized by using the same normalization method as the training data. Then, the weight coefficients of each input feature are calculated by using an attention allocator. Next, the weighted feature vector is input into a bidirectional long short-term memory network for forward and backward propagation calculation. Finally, the remaining life prediction value of the battery to be measured is obtained through a full connection layer and an output layer. The confidence interval of the prediction result is plus or minus 10%.

[0130] Specifically, the principle of the present application is: first, for the important state parameter of battery internal resistance, the present application adopts a fast voltage and current acquisition sensor, which can obtain the internal resistance data of the battery in real time and high precision. Internal resistance is a key indicator reflecting the electrochemical performance of the battery, which will increase continuously with the aging of the battery, so the internal resistance data is the key input for predicting the remaining life of the battery. In order to eliminate the interference of temperature on the measurement of internal resistance, the present application also introduces a temperature compensation model, which establishes the mapping relationship between internal resistance and temperature through heat conduction equation, temperature distribution equation, etc., and finally obtains the internal resistance value at standard temperature.

[0131] Secondly, the present application also obtains other important state parameters of the battery, such as actual discharge capacity, coulomb efficiency, charge and discharge cycle number, maximum and minimum temperature, etc., to construct a complete battery feature vector. These parameters comprehensively reflect the use state and historical operation trajectory of the battery, providing necessary input basis for subsequent remaining life prediction.

[0132] Finally, the present application adopts an improved bidirectional long short-term memory (Bi-LSTM) neural network as the battery life prediction model. The network adds an attention allocator module based on the standard Bi-LSTM structure, which can automatically identify the importance of each battery feature parameter to the remaining life prediction result and assign different weights according to the importance. This weighting mechanism can highlight key parameters and improve the accuracy and generalization ability of the prediction model. Through the training of a large number of historical data of sample batteries, the prediction model can finally accurately predict the remaining service life of the battery to be tested.

[0133] The following provides an embodiment of a specific application scenario of the present application: 200 solar street lamps are installed on a certain section of road, each equipped with a solar panel and a lead-acid battery. The section is located in the suburbs and has good lighting conditions, but due to environmental factors, the service life of the batteries is generally short, averaging only 6-7 years. The operation and maintenance department hopes to accurately predict the remaining service life of each battery, so as to develop a reasonable replacement plan and reduce maintenance costs. The operation and maintenance personnel decide to use the solar street lamp battery remaining life detection method proposed by the present application for pilot application. First, the operation and maintenance personnel select 100 out of the 200 street lamps as samples and conduct detailed parameter acquisition and performance testing on the 100 batteries. The specific steps are as follows:

[0134] 1. Obtain the basic parameters of the sample battery

[0135] For the 100 sample batteries, the operation and maintenance personnel recorded the basic information such as rated capacity, working voltage, design life, and internal structure parameters. Taking a certain battery as an example, its basic parameters are shown in Table 1:

[0136] Table 1 Basic parameters of a certain sample battery

[0137]

[0138] 2. Collecting real-time running parameters of sample batteries

[0139] In the next 3 months, the operation and maintenance personnel monitored 100 sample batteries through the signal collection device every day, and recorded the real-time voltage, real-time current, environmental temperature, battery surface temperature, charging time and discharging time and other parameters. Taking the monitoring data of a certain day as an example, as shown in Table 2:

[0140] Table 2 Real-time running parameters of 100 sample batteries on a certain day

[0141]

[0142] 3. Conducting battery discharge capacity test

[0143] After completing the real-time parameter collection for 3 months, the operation and maintenance personnel conducted a standard discharge capacity test on each of the 100 sample batteries. Through the test, the actual discharge capacity value of each battery under standard working conditions can be obtained. Taking a certain battery as an example, its actual discharge capacity is 75 Ah.

[0144] 4. Measuring battery internal resistance and conducting temperature compensation

[0145] On the basis of real-time parameter collection, the operation and maintenance personnel also measured the internal resistance of the 100 sample batteries. The specific method is to continuously record the real-time voltage value and real-time current value of each battery within a preset sampling period (such as 1 hour), and then calculate its internal resistance value :

[0146] ;

[0147] Since the internal resistance value will fluctuate with temperature changes, temperature compensation is also required. According to the temperature compensation equation group mentioned earlier, the operation and maintenance personnel calculated the internal resistance compensation value at the standard temperature :

[0148] ;

[0149] wherein is the measured internal resistance value, and are temperature coefficients, is the current temperature, is the reference temperature, and are the correction functions of state of charge and cycle number respectively. Through this step of temperature compensation, the interference of environmental temperature on internal resistance measurement can be eliminated, and more accurate internal resistance data can be obtained. Figure 3The variation curves of the battery internal resistance with the number of charge-discharge cycles at different working temperatures (10°C, 25°C and 40°C) are shown. As can be seen from the figure, with the increase of the number of charge-discharge cycles, the battery internal resistance shows a nonlinear upward trend, and the higher the temperature, the faster the internal resistance rises. The figure reflects the influence law of temperature on the change of battery internal resistance, and provides a basis for temperature compensation.

[0150] 5. Obtain the operation history data of the sample batteries

[0151] In addition to the real-time parameters, the operation personnel also collected the historical operation data of 100 sample batteries, including the number of charge-discharge cycles, cumulative charge capacity, cumulative discharge capacity, historical maximum temperature and historical minimum temperature, etc. These data reflect the use state and environmental load of the battery, which will provide an important basis for subsequent life prediction.

[0152] 6. Calculate the battery coulomb efficiency

[0153] Based on the obtained battery charge-discharge history data, the operation personnel calculated the coulomb efficiency of each sample battery :

[0154] ;

[0155] Among them, is the discharge current, is the charge current, to is the discharge time period, to is the charge time period. The coulomb efficiency reflects the energy loss of the battery during the charge-discharge process, and is an important indicator to judge the performance of the battery.

[0156] 7. Construct the battery feature vector

[0157] The above obtained various battery parameters, including the internal resistance compensation value , the coulomb efficiency , the actual discharge capacity , the number of charge-discharge cycles , the state of health , the historical maximum temperature , the historical minimum temperature and the discharge depth , are combined into a feature vector :

[0158] ;

[0159] This feature vector comprehensively describes the internal state and use environment of the battery, and provides the necessary input for subsequent life prediction.

[0160] 8. Training a battery life prediction model

[0161] Based on the feature vectors of the above-mentioned 100 sample batteries and the actual remaining life values, an improved bidirectional long short-term memory (Bi-LSTM) neural network model is established by the operation and maintenance personnel for predicting the remaining useful life of the battery. The structure of the model is as follows:

[0162] The input layer receives the feature vectors , which are weighted and processed by the attention allocator and then enter the Bi-LSTM layer. The Bi-LSTM layer includes forward-propagating LSTM units and backward-propagating LSTM units, with 64 neurons set for each LSTM unit. The output of the Bi-LSTM layer is fused through a fully connected layer, which sets 128 neurons and uses batch normalization technology to reduce data distribution bias. Finally, the battery remaining life prediction value is obtained through the linear output layer.

[0163] The attention allocator calculates the attention weight of each feature parameter through a 3-layer feedforward neural network :

[0164] ;

[0165] ;

[0166] where is the attention vector, is the weight matrix, is the i-th input feature, is the bias term, is the number of features. This weighting mechanism can highlight key parameters that have a greater impact on the remaining life prediction result.

[0167] After training a large amount of sample data, the Bi-LSTM prediction model finally achieves satisfactory prediction accuracy.

[0168] 9. Rapid detection of the battery to be tested

[0169] With the above trained battery life prediction model, the operation and maintenance personnel can quickly detect the batteries on the remaining 100 streetlights. The specific steps are as follows:

[0170] First, the key parameters of the battery to be tested are collected in real time by a signal acquisition device (as shown in Figure 2 ), including real-time voltage, real-time current, and surface temperature. The signal acquisition device is composed of 3 sensor units, 1 signal conditioning unit, and 1 control processing unit, which can quickly and accurately collect the required data.

[0171] Then, the operation and maintenance personnel calculates the internal resistance estimation value of the battery to be tested :

[0172] ;

[0173] and obtains the cumulative running time value.

[0174] Finally, the internal resistance estimation value , the surface temperature and the cumulative running time are input into the trained Bi-LSTM prediction model, and the remaining life prediction value of the battery to be tested can be obtained.

[0175] Figure 4 The performance evaluation results of the battery remaining life prediction model are shown. In the figure, the horizontal axis is the actual remaining life, the vertical axis is the predicted remaining life, the scatter points represent the corresponding relationship between the predicted value and the actual value, and the red dotted line is the ideal prediction line (the predicted value is equal to the actual value). From the distribution of the scatter points, it can be seen that the correlation between the predicted value and the actual value is high, and the prediction error is within an acceptable range.

[0176] 10. Formulate battery replacement plan

[0177] Based on the remaining life prediction results of all 100 sample batteries and 100 batteries to be tested obtained in the above steps, the operation and maintenance personnel formulates the overall battery replacement plan:

[0178] Firstly, for those batteries with a predicted remaining life of less than 2 years, immediate replacement is arranged. This part of the battery has entered the end of the aging period, and if it continues to be used, there will be a high risk of failure.

[0179] Secondly, for the batteries with a predicted remaining life of 2-5 years, the replacement time is appropriately delayed, but the inspection and monitoring efforts are increased, and the use state changes are monitored at any time.

[0180] Finally, for the batteries with a predicted remaining life of more than 5 years, they can be used according to the original plan and do not need to be replaced immediately. However, the performance indicators of the batteries should be detected regularly to discover the aging signs in time and adjust the replacement plan reasonably.

[0181] Through this differentiated replacement strategy, not only the service life of the battery can be maximized, but also the maintenance cost can be effectively controlled. It is expected that after one year of pilot application, the average service life of the solar street lamp battery of this section will be increased from 6-7 years to 8-9 years, greatly reducing the battery replacement frequency and maintenance workload.

[0182] It should be noted that the variables involved in the present application are explained as shown in Table 3.

[0183] Table 3 Variable explanation table

[0184]

[0185] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.

Claims

1. A method for detecting the remaining life of a solar street light cell, characterized by, The method comprises the following steps: S01, acquiring the basic parameters of the sample solar street lamp battery, including the battery rated capacity value, the battery operating voltage value, the battery design life value, and the battery internal structure parameter value; S02, collecting the real-time operating parameters of the sample solar street lamp battery, including the battery real-time voltage value, the battery real-time current value, the battery ambient temperature value, the battery surface temperature value, the battery charging time value, and the battery discharging time value; S03, performing battery discharge capacity testing to measure the actual discharge capacity value of the sample solar street lamp battery under standard working conditions; S04, continuously acquiring the ratio of the battery real-time voltage value to the battery real-time current value within a preset sampling period to obtain a battery internal resistance measurement sequence; S05, compensating the battery internal resistance measurement sequence based on a temperature compensation equation set to obtain a battery internal resistance compensation value at a standard temperature; S06, acquiring the operating history data of the sample solar street lamp battery, including the battery charge-discharge cycle number value, the battery cumulative charge value, the battery cumulative discharge value, the battery historical maximum temperature value, and the battery historical minimum temperature value; S07, calculating the battery coulomb efficiency value according to the battery cumulative charge value and the battery cumulative discharge value; S08, constructing a battery feature vector based on the battery internal resistance compensation value, the battery coulomb efficiency value, the actual discharge capacity value, and the battery charge-discharge cycle number value; S09, establishing a battery life neural network prediction model, inputting the battery feature vector and the corresponding actual remaining life value into the battery life neural network prediction model for training to obtain a trained battery life prediction model; S10, collecting the rapid detection parameters of the to-be-tested solar street lamp battery, including the to-be-tested battery real-time voltage value, the to-be-tested battery real-time current value, and the to-be-tested battery surface temperature value; S11, calculating the ratio of the to-be-tested battery real-time voltage value to the to-be-tested battery real-time current value to obtain a to-be-tested battery internal resistance estimation value; S12, acquiring the cumulative operating time value of the to-be-tested solar street lamp battery; S13, inputting the to-be-tested battery internal resistance estimation value, the to-be-tested battery surface temperature value, and the cumulative operating time value into the trained battery life prediction model to obtain a to-be-tested battery remaining life prediction value; The temperature compensation equation set includes a heat conduction equation, a temperature distribution equation, an internal resistance compensation equation, a heat balance equation, and a time response equation; the heat conduction equation is used to calculate the temperature conduction relationship between the inside and the surface of the battery, and the input includes the battery surface temperature value, the battery ambient temperature value, the battery internal structure parameter value, and the battery real-time current value, and the output is the battery internal temperature gradient value; the temperature distribution equation is used to describe the temperature distribution characteristics of each layer structure of the battery, and the input includes the battery internal temperature gradient value, the battery internal structure parameter value, the battery real-time current value, the battery charging time value, and the battery discharging time value, and the output is the battery temperature distribution data; the internal resistance compensation equation is used to establish the mapping relationship between the internal resistance and the temperature, and the input includes the battery temperature distribution data, the battery internal resistance measurement sequence, the battery historical maximum temperature value, and the battery historical minimum temperature value, and the output is the battery internal resistance temperature compensation coefficient value; the heat balance equation is used to calculate the steady-state working temperature of the battery, and the input includes the battery ambient temperature value, the battery real-time current value, the battery real-time voltage value, and the battery surface temperature value, and the output is the battery balance temperature value; and the time response equation is used to describe the dynamic characteristics of temperature change, and the input includes the battery balance temperature value, the preset sampling period, the battery real-time current value, and the battery real-time voltage value, and the output is the battery dynamic temperature response data.

2. The method for detecting the remaining life of a solar street lamp battery according to claim 1, characterized in that, The sample solar street lamp includes at least 100 solar street lamps. 3.The method of claim 2, wherein The battery life neural network prediction model adopts an improved bidirectional long short-term memory network structure, specifically, an attention allocator is added between a memory cell module and a forgetting gate module in the bidirectional long short-term memory network structure, the attention allocator is used to assign different weight coefficients to different feature parameters in an input battery feature vector, thereby highlighting key parameters that have a greater impact on the remaining life prediction result; the specific structure of the attention allocator is composed of a feedforward neural network, the feedforward neural network includes three hidden layers, 32 neurons are arranged in the first hidden layer, 16 neurons are arranged in the second hidden layer, and 8 neurons are arranged in the third hidden layer, the attention allocator calculates an attention weight value for each feature parameter in the input battery feature vector, and multiplies the calculated attention weight value by the original feature parameter to generate a weighted feature vector. 4.The method of claim 3, wherein The improved bidirectional long short-term memory network structure is described in detail as follows: an input layer receives the battery feature vector, and after weighted processing by the attention allocator, inputs into a bidirectional long short-term memory layer; the bidirectional long short-term memory layer includes a forward-propagating long short-term memory unit and a backward-propagating long short-term memory unit, each of which is provided with 64 neurons; the outputs of the forward-propagating long short-term memory unit and the backward-propagating long short-term memory unit are fused through a fully connected layer provided with 128 neurons; batch normalization processing is adopted to reduce the deviation of data distribution; finally, a battery remaining life prediction value is obtained through an output layer adopting a linear activation function.

Images