Remote monitoring operation and maintenance management device for energy storage lithium battery

Abstract

The invention relates to the technical field of energy storage lithium batteries, and discloses a remote monitoring operation and maintenance management device for energy storage lithium batteries, which is characterized in that a battery temperature acquisition module, a battery voltage acquisition module, a battery current acquisition module and an environment temperature acquisition module are respectively used for acquiring the surface temperature, voltage and current of a battery cell of an energy storage lithium battery group and the environment temperature of the battery; the state of charge calculation module calculates the state of charge (SOC) of the energy storage lithium battery group; the battery fault diagnosis module performs battery abnormity detection to identify a fault battery; and the early warning module generates alarm information when the SOC is not in a preset range and detects that a fault battery exists in the energy storage lithium battery group. According to the method, more data are collected through the intelligent sensor, multiple calculation modes are established to calculate the SOC of the battery, the calculation result can accurately reflect the actual working state of the energy storage power station, real-time fault detection is carried out on the energy storage lithium battery, and the use safety of the energy storage lithium battery is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage lithium batteries, and particularly relates to a remote monitoring, operation and maintenance management device for energy storage lithium batteries. Background Art

[0002] A lithium battery is a primary battery with a lithium metal or lithium alloy as the negative electrode material and a non-aqueous electrolyte solution. It is different from rechargeable lithium-ion batteries and lithium-ion polymer batteries. The inventor of the lithium battery is Edison. Due to the very active chemical properties of lithium metal, the processing, storage and use of lithium metal have very high requirements for the environment. Therefore, lithium batteries have not been applied for a long time. With the development of microelectronics technology at the end of the 20th century, the number of miniaturized devices has increased day by day, posing high requirements for power sources. As a result, lithium batteries have entered a large-scale practical stage. With the increasing upsurge of Industry 4.0, intelligent manufacturing, and Internet +, in order to accelerate the pace towards a smart factory while improving the existing production automation level, the manufacturing industry needs a more intelligent and open information system to meet the increasing production, quality, and personalized product requirements. With the development of technology, customers urgently need a product that can meet remote debugging and maintenance, perform data analysis while collecting data in real time, and can meet the needs of different departments in the same organizational system as well as the common use of different organizational systems.

[0003] At present, the SOC of lithium batteries can only be indirectly obtained through the battery management system (BMS) of the energy storage power station in combination with the SOC estimation algorithm during the actual operation of the energy storage power station. Domestic and foreign scholars have summarized the existing SOC estimation methods, but most of them are based on the models of power batteries and are therefore only applicable to small-scale energy storage systems. However, in actual applications, energy storage power stations with a large capacity not only have a large management difficulty, complex battery operating conditions, but also have frequent switching of power output and input, and most of them operate all day long with intensive working hours. Therefore, the algorithm for estimating the SOC of the energy storage power station cannot accurately reflect the complex conditions of the battery pack. At the same time, the battery failure situation also greatly affects the operation safety of the energy storage power station, and remote monitoring, operation and maintenance management of the energy storage lithium battery are required. Summary of the Invention

[0004] The present invention provides a remote monitoring, operation and maintenance management device for energy storage lithium batteries, which collects more data through intelligent sensors, establishes various calculation methods to calculate the SOC of the battery, enables the calculation results to accurately reflect the actual working state of the energy storage power station, and performs real-time fault detection on the energy storage lithium battery to ensure the safe use of the energy storage lithium battery.

[0005] The present invention provides a remote monitoring, operation and maintenance management device for energy storage lithium batteries, which includes a microprocessor, a battery temperature acquisition module, a battery voltage acquisition module, a battery current acquisition module, an ambient temperature acquisition module, a state of charge calculation module, a battery fault diagnosis module and an early warning module. The microprocessor is respectively connected to the battery temperature acquisition module, the battery voltage acquisition module, the battery current acquisition module, the ambient temperature acquisition module, the state of charge calculation module, the battery fault diagnosis module and the early warning module;

[0006] The battery temperature acquisition module, the battery voltage acquisition module, the battery current acquisition module and the ambient temperature acquisition module are respectively used to acquire the surface temperature, voltage, current of the battery cells in the energy storage lithium battery group and the ambient temperature where the battery is located, and transmit the acquired surface temperature, voltage, current of the battery cells and the ambient temperature data of the battery to the microprocessor;

[0007] The state of charge calculation module is used to calculate the state of charge SOC of the energy storage lithium battery group according to the surface temperature, voltage, current of the battery cells and the ambient temperature data of the battery; the battery fault diagnosis module is used to perform battery anomaly detection according to the voltage change data of the energy storage lithium battery to identify the faulty batteries in the energy storage lithium battery group;

[0008] The early warning module is used to generate an alarm message, perform an audible and visual alarm process, and transmit the alarm message when the state of charge SOC of the energy storage lithium battery group is not within the preset range and when a faulty battery is detected in the energy storage lithium battery group.

[0009] Further, in the data acquisition of the energy storage lithium battery group, LTC6804 is used as the battery voltage acquisition module to acquire the voltage of the energy storage lithium battery group, a Hall current sensor is used as the battery current acquisition module to acquire the current of the energy storage lithium battery group, an NTC thermistor is used as the battery temperature acquisition module to acquire the surface temperature of the battery cells, and a humidity sensor LOR is used as the ambient temperature acquisition module to acquire the humidity of the environment where the battery is located.

[0010] Further, the steps of the state of charge calculation module for calculating the state of charge SOC of the energy storage lithium battery group include:

[0011] S101. Construct an OCV-SOC curve of the energy storage lithium battery, and divide the working state interval of the energy storage lithium battery according to the OCV-SOC curve; wherein, the working state interval includes a calibration and identification interval and a voltage stability interval;

[0012] S102. Obtain the initial SOC value and the OCV value of the battery voltage of the energy storage lithium battery, and determine whether the energy storage lithium battery is overcharged or over-discharged;

[0013] S103. When the energy storage lithium battery is overcharged or overdischarged, directly search for data through the database of the OCV-SOC curve to calculate the current SOC value;

[0014] S104. When the energy storage lithium battery is not overcharged or overdischarged, determine the current working state interval of the energy storage lithium battery according to the OCV-SOC curve and the battery voltage OCV value;

[0015] S105. When the energy storage lithium battery is in the calibration and identification interval, calculate the current SOC value through a preset dU / dQ characteristic curve or Kalman filtering algorithm;

[0016] S106. When the energy storage lithium battery is in the voltage stable interval, calculate the current SOC value by using a preset ampere-hour integration method.

[0017] Further, the step S101 specifically includes:

[0018] Charge and discharge the energy storage lithium battery by using constant current-constant voltage, and calculate the amount of electricity ΔQ charged and discharged by the energy storage lithium battery with current i in the time period from t1 to t2 as:

[0019]

[0020] When the amount of electricity of the energy storage lithium battery changes constantly during charge and discharge, obtain the voltage change rate ΔU / Δt and the electricity change rate ΔQ / Δt of the entire charge and discharge process, and obtain the real-time ΔU / ΔQ curve of the entire charge and discharge process according to the voltage change rate ΔU / Δt and the electricity change rate ΔQ / Δt, and determine the voltage change rate of the energy storage lithium battery at different amounts of electricity;

[0021] When charging with constant current, then:

[0022]

[0023] ΔU / ΔQ is:

[0024]

[0025] Among them, dU / dQ is a function with voltage as a variable. In the case of constant current, dU / dQ represents the voltage change of the energy storage lithium battery within a set time period, and the SOC of the energy storage lithium battery is fed back through the voltage change of the energy storage lithium battery within the set time period;

[0026] When the energy storage lithium battery is charging, when the monomer voltage reaches 3.6V, the battery SOC is 100%, that is, it is fully charged; when the energy storage lithium battery is discharging, when the monomer voltage reaches 2.0, the battery SOC is 0%, that is, it is fully discharged; set the range beyond these two voltages as overcharge and overdischarge.

[0027] Further, in the step S106, the calculation formula is as follows:

[0028]

[0029] where SOC0 is the initial SOC value of the energy storage lithium battery, C E is the rated capacity of the energy storage lithium battery, η is the charge-discharge efficiency coefficient, that is, the Coulomb efficiency coefficient, which represents the power dissipation inside the battery during the charge-discharge process, mainly based on the charge-discharge rate and the temperature correction coefficient; I(t) is the charge-discharge current of the battery at time t, and t is the charge-discharge time.

[0030] Further, the steps for the battery fault diagnosis module to perform battery anomaly detection based on the voltage change data of the energy storage lithium battery include:

[0031] S201. Obtain the voltage data of the energy storage lithium battery group;

[0032] S202. Calculate the voltage MAD eigenvalue of the energy storage lithium battery group, and use NLM to filter the MAD eigenvalue to clearly amplify the difference between the faulty battery and the normal battery;

[0033] S203. Extract the features of the MAD value after NLM filtering as the input of the DTW algorithm to calculate the DTW similarity distance between the eigenvalue curve of each single battery and the reference single battery eigenvalue curve;

[0034] S204. Set the adaptive threshold of the corresponding DTW distance. When the DTW similarity distance is not within the range of the adaptive threshold, it is determined that the single battery has a fault.

[0035] Further, the step S202 specifically includes:

[0036] Set the voltage matrix composed of the historical data of the energy storage lithium battery group as:

[0037]

[0038] where M and N are the length of the sampling points and the number of single batteries respectively, and U M,N is the matrix composed of multiple voltage data of the energy storage lithium battery group, and u M,N is the voltage of the Nth single battery at the Mth moment of the battery;

[0039] Extract the median voltage in the voltage matrix U M,N , that is, extract the median of the voltages of all single batteries in the voltage matrix, denoted as u median , calculate the MAD value M ADk of the voltage of each single battery in the voltage matrix and the median voltage within time k, and draw the corresponding MAD eigenvalue curve:

[0040] M ADk = median(|U M,N - U median |) k

[0041] Furthermore, a new voltage eigenvalue matrix is obtained Expressed as: Wherein, is the fluctuation eigenvector of the battery voltage at time t;

[0042] Let f(t) be the MAD signal of the noise-free true voltage sequence, and θ(t) be the background noise signal. The actual noisy voltage MAD signal s(t) is expressed as: s(t) = f(t) + θ(t);

[0043] The NLM algorithm calculates the weighted average of all similar blocks in s(t) Estimate the original f(t), The calculation process of is:

[0044]

[0045] Wherein, z(t) = ∑ t∈M(t) w(t, s) is a normalized constant, and the weight w(t, s) satisfies 0 < w(t, s) < 1, and ∑ t∈M(t) w(t, s) indicates that it compares the neighborhood similarity around samples t and s;

[0046] In w(t, s), if the neighborhood of sample t is similar to the neighborhood of sample s, the weight takes the maximum value, otherwise it takes the minimum value. Its calculation method is:

[0047]

[0048] Wherein, δ is a filtering parameter for controlling the smoothing degree, λ is the step size of the filter, B Δ is the total sample, Δ represents the similar block of sample t, and changes around sample t and B Δ the total sample, d 2 is the sum of the squares of the point-to-point differences of the samples in the patch centered on s, and the value is ∑ λ∈Δ [f(t + λ) - f(s + λ)] 2 .

[0049] Furthermore, the step S203 specifically includes:

[0050] Set the sequences of the MAD eigenvalue curves of the voltages of two single cells as: X = (x1, x2,..., x a ) and Y = (y1, y2,..., ya ) There are many paths with a sequence length of a that satisfy the boundary, monotonicity, and constraint conditions, forming a path set W, which is expressed as:

[0051]

[0052] Among them, represents x in the X sequence h corresponding to y in the Y sequence h All paths w together form the combination W; the shortest path in the path sets W of the X sequence and the Y sequence is denoted as DTW(X,Y). To calculate DTW(X,Y), the Euclidean distance dis(x q , y p ) is calculated:

[0053]

[0054] The DTW calculation is from the 1st to the qth element of the sequence, and the method is:

[0055]

[0056] Calculate the DTW distance to measure the similarity between the battery voltage characteristic sequences. The larger the DTW distance, the smaller the similarity; using the definition of the DTW distance, calculate the DTW distance D between the reference battery characteristic unit X1 and the to-be-tested single battery characteristic unit X2 S : D S =(X1,X2).

[0057] Furthermore, in the step S204, the adaptive threshold T is set as: T = 1.2×[mean(D S ) + 3×std(D S )], where mean is the average value of the DTW distances of all single batteries DS at the current moment, and std is the standard deviation of the DTW distances of all single batteries DS at the current moment.

[0058] Furthermore, it further includes a cloud server and a user terminal. The cloud server is respectively connected to the microprocessor and the user terminal. The cloud server is used to receive the state of charge calculation results, fault diagnosis results, and alarm information of the energy storage lithium battery for storage and display. The user terminal is used to obtain the state of charge calculation results, fault diagnosis results, and alarm information of the energy storage lithium battery from the cloud server to achieve remote monitoring and operation and maintenance management. The user terminal includes a computer, a tablet, and a mobile phone.

[0059] The beneficial effects of the present invention are:

[0060] 1. The present invention uses a battery temperature acquisition module, a battery voltage acquisition module, a battery current acquisition module, and an ambient temperature acquisition module to acquire the surface temperature, voltage, current of the battery cells in the energy storage lithium battery group and the ambient temperature where the battery is located, and divides the working state intervals of multiple energy storage lithium batteries to establish multiple calculation methods to calculate the SOC value of the battery according to the acquired data, so that the calculation result can accurately reflect the actual working state of the energy storage power station.

[0061] 2. The present invention performs abnormal detection on the energy storage lithium battery based on the voltage curve, amplifies the voltage curve characteristics of the faulty unit through the Median Absolute Deviation (MAD), obtains obvious fault characteristics by using the Non-Local Means (NLM) filtering algorithm, calculates the similarity distance of the eigenvalue curve using the Dynamic Time Warping (DTW) algorithm, and automatically locates the faulty battery in combination with a preset adaptive threshold to ensure the safe use of the energy storage lithium battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a schematic structural diagram of the remote monitoring, operation and maintenance management device for the energy storage lithium battery of the present invention.

[0063] The realization, functional features and advantages of the object of the present invention will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0064] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0065] As Figure 1 shown, the present invention provides a remote monitoring, operation and maintenance management device for an energy storage lithium battery, including a microprocessor, a battery temperature acquisition module, a battery voltage acquisition module, a battery current acquisition module, an ambient temperature acquisition module, a state of charge calculation module, a battery fault diagnosis module, an early warning module, a cloud server and a user terminal. The microprocessor is respectively connected to the battery temperature acquisition module, the battery voltage acquisition module, the battery current acquisition module, the ambient temperature acquisition module, the state of charge calculation module, the battery fault diagnosis module and the early warning module. The cloud server is respectively connected to the microprocessor and the user terminal.

[0066] The battery temperature acquisition module, battery voltage acquisition module, battery current acquisition module, and ambient temperature acquisition module are respectively used to acquire the surface temperature, voltage, current of the energy storage lithium battery group, and the ambient temperature where the battery is located, and transmit the acquired surface temperature, voltage, current, and ambient temperature data of the battery to the microprocessor; among them, LTC6804 is used as the battery voltage acquisition module to acquire the voltage of the energy storage lithium battery group, a Hall current sensor is used as the battery current acquisition module to acquire the current of the energy storage lithium battery group, an NTC thermistor is used as the battery temperature acquisition module to acquire the surface temperature of the battery cells, and a humidity sensor LOR is used as the ambient temperature acquisition module to acquire the humidity of the environment where the battery is located.

[0067] The state of charge calculation module is used to calculate the state of charge SOC of the energy storage lithium battery group according to the surface temperature, voltage, current, and ambient temperature data of the battery cells; the battery fault diagnosis module is used to perform battery anomaly detection according to the voltage change data of the energy storage lithium battery to identify the faulty batteries in the energy storage lithium battery group;

[0068] The warning module is used to generate an alarm message, perform audible and visual alarm processing, and transmit the alarm message to the cloud server when the state of charge SOC of the energy storage lithium battery group is not within the preset range and when a faulty battery is detected in the energy storage lithium battery group. The cloud server is used to receive the state of charge calculation result, fault diagnosis result, and alarm message of the energy storage lithium battery for storage and display. The user terminal is used to obtain the state of charge calculation result, fault diagnosis result, and alarm message of the energy storage lithium battery from the cloud server to achieve remote monitoring and operation and maintenance management. The user terminal includes a computer, a tablet, and a mobile phone.

[0069] Specifically, in the state of charge calculation module, the steps of calculating the state of charge SOC of the energy storage lithium battery group by the calculation module include:

[0070] S101. Construct an OCV-SOC curve of the energy storage lithium battery, and divide the working state interval of the energy storage lithium battery according to the OCV-SOC curve; among them, the working state interval includes a calibration identification interval and a voltage stable interval.

[0071] The energy storage lithium battery is charged and discharged by constant current-constant voltage. In this way, the potential of the energy storage lithium battery always changes at a constant rate during the charge and discharge process. Calculate the amount of charge ΔQ charged and discharged by the energy storage lithium battery with current i in the time period from t1 to t2 as:

[0072]

[0073] Keep the power of the energy storage lithium battery changing constantly during charging and discharging, obtain the voltage change rate ΔU / Δt and the charge change rate ΔQ / Δt of the entire charging and discharging process, and obtain the real-time ΔU / ΔQ curve of the entire charging and discharging process based on the voltage change rate ΔU / Δt and the charge change rate ΔQ / Δt, and determine the voltage change rate of the energy storage lithium battery at different charges;

[0074] When charging at a constant current, then:

[0075]

[0076] ΔU / ΔQ is:

[0077]

[0078] Wherein, dU / dQ is a function with voltage as a variable. Under the condition of constant current, dU / dQ represents the voltage change of the energy storage lithium battery within a set time period, and the SOC of the energy storage lithium battery is fed back through the voltage change of the energy storage lithium battery within the set time period; meanwhile, during the operation of an actual energy storage power station, the voltage of the energy storage lithium battery can be obtained in real time. Therefore, the above curve is refined and applied to the SOC calculation of the energy storage power station.

[0079] When the energy storage lithium battery is charging, when the single-cell voltage reaches 3.6V, the battery SOC is 100%, which means it is fully charged; when the energy storage lithium battery is discharging, when the single-cell voltage reaches 2.0, the battery SOC is 0%, which means it is fully discharged; set the range beyond these two voltages as overcharging and over-discharging.

[0080] According to the above-mentioned characteristic curve and OCV-SOC curve of the energy storage lithium battery, the state interval of the energy storage lithium battery can be divided. The Q-U curve is corrected by calibrating the battery voltage corresponding to the battery SOC percentage to obtain the characteristic curve of the energy storage lithium battery, that is, the OCV-SOC (open circuit voltage - remaining charge) curve. Through the curve, the two main state regions of the battery can be clearly judged, that is, the identification and correction region and the voltage stable region. Beyond these two voltage ranges are the overcharging and over-discharging regions. The two main state regions are divided based on the inflection points with sudden slope changes on the OCV-SOC curve. At this point, the slope change of the curve is very obvious.

[0081] In the voltage stable region of the energy storage lithium battery, the voltage change is small, but the SOC increases and decreases significantly. The open circuit voltage method has a large estimation error, so the ampere-hour integration method is more accurate for SOC estimation of the energy storage lithium battery in this interval; in the identification and calibration region of the energy storage lithium battery, the identification of SOC estimation is mainly based on the characteristic curve dU / dQ of the energy storage lithium battery. By comparing and analyzing the model database, a suitable model and data are selected to obtain the SOC of the battery; in the overcharge and over-discharge regions of the energy storage lithium battery, since the duration is very short and the OCV of the battery changes greatly, data can be directly searched through the OCV-SOC curve database for estimation, that is, steps S102 - S106.

[0082] S102. Obtain the initial SOC value and the battery voltage OCV value of the energy storage lithium battery, and determine whether the energy storage lithium battery is overcharged or over-discharged.

[0083] S103. When the energy storage lithium battery is overcharged or over-discharged, directly search for data through the OCV-SOC curve database to calculate the current SOC value.

[0084] S104. When the energy storage lithium battery is not overcharged or over-discharged, determine the current working state interval of the energy storage lithium battery according to the OCV-SOC curve and the battery voltage OCV value.

[0085] S105. When the energy storage lithium battery is in the calibration and identification interval, calculate the current SOC value through a preset dU / dQ characteristic curve or Kalman filter algorithm.

[0086] S106. When the energy storage lithium battery is in the voltage stable interval, calculate the current SOC value using a preset ampere-hour integration method. The calculation formula is:

[0087]

[0088] where SOC0 is the initial SOC value of the energy storage lithium battery, C E is the rated capacity of the energy storage lithium battery, η is the charge and discharge efficiency coefficient, that is, the Coulomb efficiency coefficient, which represents the power dissipation inside the battery during the charge and discharge process, mainly based on the charge and discharge rate and the temperature correction coefficient; I(t) is the charge and discharge current of the battery at time t, and t is the charge and discharge time.

[0089] In the battery fault diagnosis module, the steps for battery anomaly detection based on the voltage change data of the energy storage lithium battery include:

[0090] S201. Obtain the voltage data of the energy storage lithium battery group.

[0091] S202. Calculate the voltage MAD eigenvalue of the energy storage lithium battery group, and use NLM to filter the MAD eigenvalue to clearly amplify the difference between the faulty battery and the normal battery.

[0092] MAD is a robust measurement of the sample deviation of data for a variable, used to describe the standard by which a single-variable sample varies in quantitative data. When there is an early fault in the energy storage lithium battery pack, the faulty batteries are in the minority and the vast majority of the batteries are normal. Therefore, the median of each sample can reflect the overall trend and is considered the data of the normal batteries. Set the voltage matrix composed of the historical data of the energy storage lithium battery group as:

[0093]

[0094] where M and N are the length of the sampling points and the number of single batteries respectively, and U M,N is the matrix composed of multiple voltage data of the energy storage lithium battery group, and u M,N is the voltage of the Nth single battery at the Mth moment of the battery;

[0095] Extract the median voltage in the voltage matrix U M,N , that is, extract the median of the voltages of all single batteries in the voltage matrix, denoted as u median , calculate the MAD value M ADk of the voltage of each single battery in the voltage matrix and the median voltage within time k, and draw the corresponding MAD eigenvalue curve:

[0096] M ADk = median(|U M,N - u median |) k

[0097] Furthermore, obtain a new voltage eigenvalue matrix expressed as:

[0098]

[0099] where is the voltage fluctuation eigenvector of the battery at time t, and the voltage MAD value of the normal unit is close to 0.

[0100] The calculation process of MAD will be interfered by the random perturbation of voltage noise. Therefore, to reduce the interference caused by the noise signal, the NLM filtering algorithm can be used to smooth the initial MAD eigenvalue curve. Let f(t) be the MAD signal of the noiseless true voltage sequence, and θ(t) be the background noise signal. The actual noisy voltage MAD signal s(t) is expressed as: s(t) = f(t) + θ(t);

[0101] The algorithm based on non - local means denoising is a process of recovering the true signal f(t) from the actual signal s(t) containing noise θ(t) by eliminating noise. The NLM algorithm calculates the weighted average of all similar blocks in s(t). Estimate the original f(t) to achieve noise reduction. The calculation process is as follows:

[0102]

[0103] Among them, z(t)=∑ t∈M(t) w(t, s) is a normalized constant, and the weight w(t, s) satisfies 0 < w(t, s) < 1, and ∑ t∈M(t) w(t, s) indicates that it compares the neighborhood similarity around samples t and s.

[0104] In w(t, s), if the neighborhood of sample t is similar to the neighborhood of sample s, the weight takes the maximum value, otherwise it takes the minimum value. Its calculation method is:

[0105]

[0106] Among them, δ is a filtering parameter that controls the smoothness degree, λ is the step size of the filter, B Δ is the total sample, Δ represents the similar block of sample t, which changes around sample t and the total sample B Δ d 2 is the sum of the squares of the point - to - point differences of the samples in the patch centered on s, and its value is ∑ λ∈Δ [f(t + λ)-f(s + λ)] 2 .

[0107] If δ in NLM is set too small, it may lead to an increase in noise fluctuations, resulting in insufficient averaging between similar blocks. If it is set too large, the filtered signal is too smooth, resulting in serious loss of details. δ is proportional to the noise standard deviation σ. When δ = σ, the effect is better.

[0108] S203. The features after the extracted MAD values are filtered by NLM are used as the input of the DTW algorithm to calculate the DTW similarity distance between the eigenvalue curves of each single - cell battery and the reference single - cell battery eigenvalue curve.

[0109] After the battery voltage data of the energy - storage lithium - battery group is extracted by MAD and NLM features, the voltage characteristics of the faulty battery in the battery pack can be amplified. Since the actual number of batteries in the energy - storage lithium - battery group is very large, resulting in a large amount of data, it is necessary to automatically detect and locate the faulty battery.

[0110] The DTW algorithm is an algorithm used to measure the similarity between two time series. It uses dynamic programming under certain constraints to solve for the cumulative minimum distance of the match and describe the similarity, and has a better measurement effect on data reconstruction. Therefore, the DTW algorithm is used to automatically locate and identify faulty batteries, specifically including:

[0111] Set the sequences of the MAD eigenvalue curves of the voltages of two single batteries as: X = (x1, x2, …, x a ) and Y = (y1, y2, …, y a ), the sequence length is a, and there are many paths that satisfy the boundary, monotonicity, and constraint conditions, forming a path set W, expressed as:

[0112]

[0113] Among them, represents the correspondence between x h in the X sequence and y h in the Y sequence. All paths w together form the combination W; the shortest path in the path sets W of the X sequence and the Y sequence is denoted as DTW(X, Y). To calculate DTW(X, Y), the Euclidean distance dis(x q , y p ) is calculated:

[0114]

[0115] The DTW calculation is from the 1st to the qth element of the sequence, and the method is:

[0116]

[0117] Calculate the DTW distance to measure the similarity between the battery voltage characteristic sequences. The larger the DTW distance, the smaller the similarity; using the definition of the DTW distance, calculate the DTW distance D S between the reference battery characteristic unit X1 and the characteristic unit X2 of the single battery to be measured: D S = (X1, X2).

[0118] When calculating the DTW distance between the characteristic values of the single battery to be measured and the reference single battery, considering that there may be abnormal batteries in the battery pack, if the average value of the voltage characteristic value curve is selected as the reference unit or the DTW distance between the voltage characteristic value curves of the batteries is calculated, it may cause the result to drift and result in misdiagnosis. The median voltage characteristic value curve can be selected as the reference characteristic value curve, and the DTW distance corresponding to each single battery is calculated, which can not only avoid misdiagnosis caused by the drift of the calculation result, but also reduce misdiagnosis caused by the inconsistency of the single batteries.

[0119] S204. Set the adaptive threshold for the corresponding DTW distance. When the DTW similarity distance is not within the range of the adaptive threshold, it is determined that the single cell has a fault.

[0120] The 3-δ criterion is a common outlier detection method and the main diagnostic standard for lithium batteries as a fault threshold. Set the adaptive threshold T for the corresponding DTW distance to determine whether the battery has a fault. If the data is not within the range of the threshold T, it will be regarded as a fault point. The adaptive threshold T is set as:

[0121] T = 1.2 × [mean(D S ) + 3 × std(D S )]

[0122] where mean is the average DTW distance of all single cells' DS at the current moment, and std is the standard deviation of the DTW distance of all single cells' DS at the current moment, that is, T is 1.2 times the 3-δ criterion.

[0123] Generally, during the discharge process of the battery, it will be affected by other factors, and the voltage will occasionally fluctuate, resulting in fluctuations in the similarity of the voltage characteristic value curve. When setting the adaptive threshold according to the 3-δ criterion, the threshold margin needs to be considered to prevent false alarms of the set adaptive threshold. Normally, there is no alarm.

[0124] The present invention uses a battery temperature acquisition module, a battery voltage acquisition module, a battery current acquisition module, and an ambient temperature acquisition module to acquire the surface temperature, voltage, current of the battery cells in the energy storage lithium battery group, and the ambient temperature where the battery is located, and divides the working state intervals of multiple energy storage lithium batteries to establish multiple calculation methods to calculate the SOC value of the battery according to the acquired data, so that the calculation result can accurately reflect the actual working state of the energy storage power station. In addition, the present invention also performs abnormal detection on the energy storage lithium battery based on the voltage curve, amplifies the voltage curve characteristics of the faulty unit through the median absolute deviation (MAD), and uses the non-local means (NLM) filtering algorithm to obtain obvious fault characteristics, calculates the similarity distance of the eigenvalue curve using the dynamic time warping (DTW) algorithm, and automatically locates the faulty battery in combination with the preset adaptive threshold to ensure the safe use of the energy storage lithium battery.

[0125] It should be noted that in this article, the terms "including", "comprising" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent to such process, device, article or method. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, device, article or method including that element.

[0126] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A remote monitoring, operation and maintenance management device for an energy storage lithium battery, characterized in that, It includes a microprocessor, a battery temperature acquisition module, a battery voltage acquisition module, a battery current acquisition module, an ambient temperature acquisition module, a state of charge calculation module, a battery fault diagnosis module, and an early warning module. The microprocessor is respectively connected to the battery temperature acquisition module, the battery voltage acquisition module, the battery current acquisition module, the ambient temperature acquisition module, the state of charge calculation module, the battery fault diagnosis module, and the early warning module; The battery temperature acquisition module, the battery voltage acquisition module, the battery current acquisition module, and the ambient temperature acquisition module are respectively used to acquire the surface temperature, voltage, current of the battery cells in the energy storage lithium battery group, and the ambient temperature where the battery is located, and transmit the acquired surface temperature, voltage, current of the battery cells, and ambient temperature data of the battery to the microprocessor; The state of charge calculation module is used to calculate the state of charge SOC of the energy storage lithium battery group according to the surface temperature, voltage, current of the battery cells, and ambient temperature data of the battery; The battery fault diagnosis module is used to perform battery anomaly detection according to the voltage change data of the energy storage lithium battery to identify the faulty batteries in the energy storage lithium battery group; The early warning module is used to generate an alarm message, perform audible and visual alarm processing, and transmit the alarm message when the state of charge SOC of the energy storage lithium battery group is not within the preset range and when a faulty battery is detected in the energy storage lithium battery group.

2. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 1, characterized in that, In the data acquisition of the energy storage lithium battery group, LTC6804 is used as the battery voltage acquisition module to acquire the voltage of the energy storage lithium battery group, a Hall current sensor is used as the battery current acquisition module to acquire the current of the energy storage lithium battery group, an NTC thermistor is used as the battery temperature acquisition module to acquire the surface temperature of the battery cells, and a humidity sensor LOR is used as the ambient temperature acquisition module to acquire the humidity of the environment where the battery is located.

3. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 1, characterized in that, The steps of the state of charge calculation module for calculating the state of charge SOC of the energy storage lithium battery group include: S101. Construct an OCV-SOC curve of the energy storage lithium battery, and divide the working state interval of the energy storage lithium battery according to the OCV-SOC curve; Among them, the working state interval includes a calibration and identification interval and a voltage stable interval; S102. Obtain the initial SOC value and the OCV value of the battery voltage of the energy storage lithium battery, and determine whether the energy storage lithium battery is overcharged or over-discharged; S103. When the energy storage lithium battery is overcharged or over-discharged, directly search for data through the database of the OCV-SOC curve to calculate the current SOC value; S104. When the energy storage lithium battery is not overcharged or over-discharged, determine the current working state interval of the energy storage lithium battery according to the OCV-SOC curve and the OCV value of the battery voltage; S105. When the energy storage lithium battery is in the calibration and identification interval, calculate the current SOC value through a preset dU / dQ characteristic curve or Kalman filter algorithm; S106. When the energy storage lithium battery is in the voltage stable interval, calculate the current SOC value by using a preset ampere-hour integration method.

4. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 3, characterized in that, The specific content of step S101 includes: The energy storage lithium battery is charged and discharged by constant current - constant voltage, and the electric quantity ΔQ charged and discharged by the energy storage lithium battery with current i in the time period from t1 to t2 is calculated as follows: When the electric quantity of the energy storage lithium battery changes constantly during charging and discharging, the voltage change rate ΔU / Δt and the electric quantity change rate ΔQ / Δt of the whole charging and discharging process are obtained, and the ΔU / ΔQ curve in real time of the whole charging and discharging process is obtained according to the voltage change rate ΔU / Δt and the electric quantity change rate ΔQ / Δt, and the voltage change rate at different electric quantities of the energy storage lithium battery is determined; When charging with constant current, then: ΔU / ΔQ is: Among them, dU / dQ is a function with voltage as a variable. In the case of constant current, dU / dQ represents the voltage change of the energy storage lithium battery in the set time period, and the SOC of the energy storage lithium battery is fed back through the voltage change of the energy storage lithium battery in the set time period; When the energy storage lithium battery is charging, when the single - cell voltage reaches 3.6V, the battery SOC is 100%, which means it is fully charged; when the energy storage lithium battery is discharging, when the single - cell voltage reaches 2.0, the battery SOC is 0%, which means it is fully discharged; the range beyond these two voltages is set as over - charge and over - discharge.

5. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 3, characterized in that, In the step S106, the calculation formula is: Among them, SOC0 is the initial SOC value of the energy storage lithium battery, C E is the rated capacity of the energy storage lithium battery, η is the charge-discharge efficiency coefficient, that is, the Coulomb efficiency coefficient, which represents the power dissipation inside the battery during the charge-discharge process, mainly based on the charge-discharge rate and the temperature correction coefficient; I(t) is the charge-discharge current of the battery at time t, and t is the charge-discharge time.

6. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 1, characterized in that, The steps for the battery fault diagnosis module to detect battery anomalies according to the voltage change data of the energy storage lithium battery include: S201. Obtain the voltage data of the energy storage lithium battery group; S202. Calculate the voltage MAD eigenvalue of the energy storage lithium battery group, and use NLM to filter the MAD eigenvalue to clearly amplify the difference between the faulty battery and the normal battery; S203. Extract the characteristics of the MAD value after NLM filtering as the input of the DTW algorithm to calculate the DTW similarity distance between the eigenvalue curve of each single - cell battery and the reference single - cell battery eigenvalue curve; S204. Set the adaptive threshold of the corresponding DTW distance. When the DTW similarity distance is not within the range of the adaptive threshold, it is determined that the single - cell battery has a fault.

7. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 6, characterized in that, The step S202 specifically includes: Set the voltage matrix composed of the historical data of the energy storage lithium battery group as: where M and N are the length of the sampling points and the number of single cells respectively, and U M,N is the matrix composed of multiple voltage data of the energy storage lithium battery group, and u M,N is the voltage of the Nth single cell at the Mth moment of the battery; Extract the voltage matrix U M,N The median voltage in it, that is, the median of the voltages of all single cells in the voltage matrix, is denoted as u median , calculate the MAD value M of the voltage of each single cell in the voltage matrix and the median voltage within time k ADk , and plot the corresponding MAD eigenvalue curve: M ADk = median(|U M,N - u median |) k Furthermore, a new voltage eigenvalue matrix is obtained. It is expressed as: Wherein, is the fluctuation feature vector of the battery voltage at time t; Let f(t) be the MAD signal of the no - noise true voltage sequence, θ(t) be the background noise signal, and the actual noisy voltage MAD signal s(t) is expressed as: s(t)=f(t)+θ(t); The NLM algorithm calculates the weighted average of all similar blocks in s(t). Estimate the original f(t). The calculation process is as follows: where z(t) = ∑ t∈M(t) w(t, s) is a normalized constant, and the weight w(t, s) satisfies 0 < w(t, s) < 1, and ∑ t∈M(t) w(t, s) represents that it compares the neighborhood similarity around samples t and s; In w(t, s), if the neighborhood of sample t is similar to the neighborhood of sample s, the weight takes the maximum value, otherwise it takes the minimum value, and its calculation method is: Among them, δ is the filtering parameter for controlling the smoothness, λ is the step size of the filter, and B Δ is the total sample, Δ represents the similar block of the t sample, and it changes around the t sample and B Δ the total sample, and d 2 is the sum of the squares of the point-to-point differences of the samples in the patch centered on s, and its value is ∑ λ∈Δ [f(t + λ) - f(s + λ)] 2 .

8. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 6, characterized in that, The step S203 specifically includes: Set the sequences of the characteristic value curves of the voltages of two single-cell batteries MAD as: X = (x1, x2, …, x a ) and Y = (y1, y2, …, y a ). There are many paths that satisfy the boundary, monotonicity, and constraint conditions, forming a path set W, which is expressed as: Among them, represents x in the X sequence h corresponding to y in the Y sequence h All paths w together form the combination W; The shortest path among the path sets W of the X sequence and the Y sequence is denoted by DTW(X, Y). To calculate DTW(X, Y), the Euclidean distance dis(x q , y p ) is calculated: The DTW calculation is from the 1st to the qth element of the sequence, and the method is: Calculate the DTW distance to measure the similarity between the battery voltage feature sequences. The greater the DTW distance, the smaller the similarity. Using the definition of the DTW distance, calculate the DTW distance D between the reference battery feature unit X1 and the battery cell under test feature unit X2 S : D S =(X1, X2).

9. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 8, characterized in that, In the step S204, the adaptive threshold T is set as: T = 1.2 × [mean(D S ) + 3 × std(D S )], where mean is the average of the DTW distances of all single cells DS at the current moment, and std is the standard deviation of the DTW distances of all single cells DS at the current moment.

10. The remote monitoring, operation and maintenance management device for the energy storage lithium battery according to claim 1, characterized in that, It also includes a cloud server and a user terminal. The cloud server is respectively connected to the microprocessor and the user terminal. The cloud server is used to receive the calculation results of the state of charge, fault diagnosis results, and alarm information of the energy storage lithium battery for storage and display. The user terminal is used to obtain the calculation results of the state of charge, fault diagnosis results, and alarm information of the energy storage lithium battery from the cloud server to realize remote monitoring, operation, and maintenance management. The user terminal includes a computer, a tablet, and a mobile phone.

Images