A lithium battery abnormality monitoring method and system based on neuron detection

Abstract

The present invention belongs to the technical field of battery safety control, and particularly relates to an abnormal monitoring method and system for lithium batteries based on neuron detection. The method uses a binary gas sensor prepared from TiO2 and SnO2 as a neuron sensor to detect different positions in the battery cabinet. The signal analysis process includes three major stages: abnormal signal capture, signal time-domain analysis, and environmental interference self-check. In the first stage, abnormal points of characteristic gases in the lithium battery cabinet are found through two indicators: the duration of concentration exceeding the limit and the rate of concentration change. In the second stage, the causal relationship of the abnormality is analyzed by combining the time-domain characteristics of different abnormal points, and then environmental interference is preliminarily excluded. In the third stage, active exhaust is used to accurately eliminate the interference of characteristic gases in the environment on the monitoring process. Finally, early detection and abnormal positioning of characteristic gases in lithium batteries are realized. The present invention overcomes the defects of poor accuracy, insufficient real-time performance, and high false alarm rate existing in the traditional safety monitoring process of lithium battery cabinets.

Description

Technical Field

[0001] The present invention belongs to the technical field of battery safety control, and specifically relates to a method for monitoring an abnormality of a lithium battery based on neuron detection, and an abnormality monitoring system for a lithium battery cabinet based on a neuron sensor array. Background Art

[0002] As human environmental awareness continues to improve, technologies such as clean energy generation and electric vehicles are also constantly developing and gradually being promoted and applied. Among them, electric vehicles need to store electricity during driving, and solar and wind power stations also need to temporarily store the generated electricity in order to achieve a balance between supply and demand of electricity. These all require high-performance energy storage carriers. On this basis, batteries, as the most convenient energy storage technology, are also expanding their market size. Among them, power lithium batteries have become the main carrier for power storage in electric vehicles and energy storage power stations due to their high energy density, high-rate charge and discharge performance, and long cycle life.

[0003] Although power lithium batteries have superior performance, they also have their own defects, namely: the safety risk of batteries is relatively high. As we all know, lithium batteries contain materials that are prone to violent chemical reactions. Therefore, lithium batteries will spontaneously combust or even explode when they are short-circuited or subjected to external impact. Moreover, the accident process is often relatively rapid, and the window time for safe disposal is relatively short. Therefore, how to detect and warn the abnormal state of lithium batteries at an early stage has become a necessary way to reduce battery safety risks.

[0004] At present, the early warning of abnormal conditions of lithium batteries mainly focuses on the analysis and warning of characteristic gases when the battery is abnormal. Taking lithium battery cabinets as an example, the existing technology usually uses sensors to obtain characteristic gas parameters in the environment, and then analyzes the monitoring data based on expert experience to predict the safety risks of lithium batteries. This type of solution has indeed achieved good monitoring results in some scenarios, which has won time for managers to detect and control accidents early. However, the above solution still has the following bottlenecks: (1) There is a lack of dedicated gas sensors for the characteristic gases generated by specific lithium batteries under abnormal conditions, which makes it difficult to meet the expected response characteristics of characteristic gases. (2) The existing solution mainly monitors the characteristic gases in the environment, cannot detect the early risks of specific battery packs, cannot achieve risk positioning, and the real-time performance of risk monitoring is relatively insufficient. (3) The early warning results of risk monitoring are easily affected by environmental factors and have a high false alarm rate. Summary of the invention

[0005] In order to solve the defects in the safety monitoring of lithium battery cabinet products, such as poor characteristic gas detection accuracy, insufficient real-time performance of risk monitoring schemes and high false alarm rate, the present invention provides a lithium battery abnormality monitoring method and system based on neuron detection.

[0006] The present invention is implemented by the following technical solutions:

[0007] The present invention provides a method for monitoring lithium battery anomalies based on neuron detection, which determines whether there is a safety risk in the lithium battery cabinet by real-time detection of the concentration of characteristic gases at different detection sites in the lithium battery cabinet. The method includes the following process:

[0008] 1. Abnormal signal capture

[0009] The real-time concentration Q of the characteristic gas at the air outlet, air inlet and detection points corresponding to each battery module of the lithium battery cabinet is collected in real time through the neural sensor, and the real-time concentration change rate V is calculated.

[0010] The real-time concentration at any detection site reaches the preset safety threshold Q max When , an over-limit signal is generated and the over-limit time t0 is recorded.

[0011] Determine whether the duration ΔT and concentration change rate of the over-limit signal at each detection site have reached their respective preset safety thresholds: if so, an abnormal state signal is generated; the abnormal state signal contains the state flag Abn and the coding information of the detection site.

[0012] 2. Signal time domain analysis

[0013] When an abnormal status signal is generated at any battery module, it is determined whether there is an abnormal status signal in the historical data at the air inlet before the preset delay period: if so, the abnormality investigation state is entered; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

[0014] When there is a safety risk of electrolyte leakage, the abnormal battery module is located according to the coded information in the abnormal status signal, and then a risk warning signal containing the location information is issued.

[0015] 3. Environmental interference self-check

[0016] In the abnormal troubleshooting state, close the air inlet and the air inlet fan, and execute the exhaust cycle of the preset duration.

[0017] Determine whether the real-time concentration Q and concentration change rate V of the characteristic gas at the battery modules and air outlets with abnormal status signals during the exhaust cycle have dropped below the safety threshold: if so, it is determined that the abnormal status signal at the battery module is caused by an abnormal external environment; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

[0018] When there is a safety risk of electrolyte leakage, the abnormal battery module is located according to the coded information in the abnormal status signal, and then a risk warning signal containing the location information is issued.

[0019] As a further improvement of the present invention, in the abnormal signal capture stage, the concentration change rate is calculated as follows:

[0020]

[0021] As a further improvement of the present invention, the process of locating the electrolyte leakage point is as follows:

[0022] In the abnormal signal capture stage, each detection site corresponds to an exclusive coding information; the generated abnormal state signal includes the state flag Abn and coding information of each detection site.

[0023] During the signal time domain analysis and environmental interference self-check stage, the preset coding rule table is queried according to the coding information in the abnormal state signal, and the location information of the abnormal battery module is decoded; a one-to-one mapping relationship between the coding information and the location information is established in the coding rule table.

[0024] As a further improvement of the present invention, in the abnormal signal capture stage, the generation formula of the state flag Abn of the abnormal state signal is as follows:

[0025]

[0026] In the above formula, V max Indicates the preset safety threshold of the characteristic gas concentration change rate; T max Indicates the preset safety threshold of the exceeding time limit.

[0027] As a further improvement of the present invention, in the signal time domain analysis stage, the preset delay period refers to the fluid delivery time from the air outlet to each battery module under the current fan speed conditions of the air inlet and the air outlet. The preset delay period of the detection point at each battery module in the lithium battery cabinet is positively correlated with the length of the air duct from the air inlet.

[0028] As a further improvement of the present invention, in the signal time domain analysis stage, when there are abnormal status signals at the detection points at multiple battery modules and the air inlet, it is determined whether the time difference between the over-limit moment at each battery module and the over-limit moment at the air inlet is positively correlated with the length of the air duct from each to the air inlet. If so, the abnormal investigation state is entered; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

[0029] As a further improvement of the present invention, in the environmental interference self-check stage, when the real-time concentration Q of the characteristic gas detected at any detection point is lower than the preset safety threshold Q max When the abnormal status signal of the current detection site is updated, the status flag is set to 0.

[0030] The present invention also includes an abnormality monitoring system for a lithium battery cabinet based on a neuron sensor array, which includes a neuron sensor array and a centralized analyzer.

[0031] The neuron sensor array includes multiple Pack gas detectors and multiple reference gas detectors. The Pack gas detectors are installed at each battery module in the lithium battery cabinet to detect the real-time concentration of the characteristic gas at the battery module. The reference gas detectors are installed at the air inlet and outlet of the battery cabinet to detect the real-time concentration of the characteristic gas at the air inlet and outlet of the lithium battery cabinet. Both the Pack gas detector and the reference gas detector use the aforementioned gas sensitive sensor.

[0032] The centralized analyzer is electrically connected to each Pack gas detector and reference gas detector in the neuron sensor array. The centralized analyzer includes a data processing unit, a positioning unit and a communication unit; the data processing unit processes the collected sensor detection data using the aforementioned abnormal monitoring method for lithium batteries based on neuron detection, and then determines whether there is a safety risk of electrolyte leakage in the lithium battery cabinet according to the real-time concentration of the characteristic gas at each detection site. The positioning unit is used to query a "code-position" comparison table to determine the spatial position of the battery module when there is a safety risk in any battery module. The communication unit is used to send out warning information representing safety risks and the location information of the battery module.

[0033] As a further improvement of the present invention, the Pack gas detector and the reference gas detector in the neuron sensor array both use a gas sensor specifically used to detect the concentration of characteristic gases containing any one or more components of dimethyl carbonate, diethyl carbonate, and ethyl methyl carbonate. The gas sensor includes a sensitive element and a signal processing circuit; the sensitive element uses a TiO 2 and SnO 2 The signal processing circuit is used to convert the response of the sensitive element into a corresponding electrical signal.

[0034] As a further improvement of the present invention, the performance indicators of the gas sensor performance used in the present invention include:

[0035] (1) The maximum sensitivity of the characteristic gas exceeds 15.

[0036] (2) The linearity within the measuring range shall not be less than 20%.

[0037] (3) Signal response time T under the lowest concentration condition res No more than 10 seconds.

[0038] (4) Signal recovery time T rec No more than 100s.

[0039] The technical solution provided by the present invention has the following beneficial effects:

[0040] The present invention proposes a method using TiO 2 and SnO 2 The gas sensor is composed of sensitive electrodes made of two semiconductor materials. The gas sensor has very excellent gas-sensitive response characteristics to the characteristic gases of dimethyl carbonate (DMC), diethyl carbonate (DEC), and ethyl methyl carbonate (EMC). The gas sensor improves the sensitivity and accuracy of lithium battery electrolyte vapor, laying the foundation for early monitoring of lithium battery abnormalities.

[0041] The present invention also uses this special gas sensor as a sensing neuron to construct an abnormal monitoring system for a lithium battery cabinet, and designs a set of special data processing and analysis logic for the monitoring system. In the monitoring scheme of the present invention, the system can conduct a comprehensive spatiotemporal analysis of detection data from different sources, discover potential abnormalities as early as possible through multi-dimensional criteria, and perform time domain analysis and abnormality investigation on abnormal information to remove detected "pseudo-abnormalities". In addition, on the basis of ensuring detection efficiency and real-time performance, the monitoring accuracy of the scheme is greatly improved, the false alarm rate is reduced, and the practical value of the scheme is ultimately enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is an electrical schematic diagram of a gas sensor provided in Example 1 of the present invention.

[0043] Figure 2 The gas sensitivity response curve of the gas sensor designed in Example 1 of the present invention to EMC steam in the concentration range of 50-2000ppm;

[0044] Figure 3 Response and recovery curves of the gas sensor in Example 1 of the present invention to 500 ppm of EMC steam.

[0045] Figure 4 This is a deployment diagram of the gas sensors in the lithium battery cabinet in the abnormal monitoring solution in Example 2 of the present invention.

[0046] Figure 5 This is a flowchart of the steps of a method for abnormality monitoring of lithium batteries based on neuron detection provided in Example 2 of the present invention.

[0047] Figure 6 This is a flow chart of the steps for locating a battery module at risk in Example 2 of the present invention.

[0048] Figure 7 This is a system architecture diagram of an abnormality monitoring system for a lithium battery cabinet based on a neural sensor array provided in Example 3 of the present invention.

[0049] Figure 8 This is a gas-sensitive response image of the neuron sensor array comprising a plurality of gas-sensitive sensors during monitoring in Example 3 of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. 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.

[0051] Example 1

[0052] This embodiment provides a gas sensor, which is specifically used to detect the concentration of characteristic gases containing any one or more components of dimethyl carbonate (DMC), diethyl carbonate (DEC), and ethyl methyl carbonate (EMC). That is, this embodiment provides an electrolyte vapor sensor. The gas sensor of this embodiment includes a sensitive element and a signal processing circuit. The sensitive element is made of TiO 2 and SnO 2 Specifically, this embodiment uses a nanowire array material made of two semiconductor oxides in a mass ratio of 1:1 as the sensitive layer. 2 O 3 As a substrate, the required resistive semiconductor gas sensor is then prepared. The signal processing circuit is used to convert the response of the sensitive element into a corresponding electrical signal.

[0053] The electrical schematic diagram of the gas sensor provided in this embodiment is as follows: Figure 1 As shown. Combined Figure 1 It can be seen that the reference resistance value of the sensitive element in the gas sensor is R 0 The steady-state resistance of the sensitive element after responding to the characteristic gas is R S ; The signal processing circuit also has a fixed resistor R in series with the sensitive element m Therefore, by measuring R m The voltage change at both ends can obtain the steady-state resistance value R of the sensitive element. S ; Steady-state resistance R S The calculation formula is as follows:

[0054]

[0055] In the above formula, V c Represents the supply voltage of the gas sensor, V m Represents the voltage division across the fixed resistor in the gas sensor.

[0056] In actual application, it is necessary to determine the response characteristics of the gas sensor through the concentration gradient test of the characteristic gas. The test process of the response characteristics is as follows: First, the tester stabilizes the gas sensor in a clean environment without characteristic gases and obtains the reference resistance value R 0 The reference resistance value R 0 After stabilization, the characteristic gas is gradually introduced into the test environment according to the preset concentration gradient. At this time, the conductivity of the sensitive element will change, and the steady-state resistance value Rs of the gas sensor will also change accordingly. By recording the steady-state resistance value of the sensitive element under different characteristic gas concentration conditions, a mapping relationship can be established between the characteristic gas concentration and the gas sensor response, and then the required gas sensitivity response curve of the gas sensor can be obtained.

[0057] The gas sensor provided in this embodiment is mainly used to detect electrolyte vapor, thereby realizing safety monitoring and early warning of the operation process of the lithium battery. Therefore, it is necessary to limit the performance indicators of the gas sensor, such as the range, maximum sensitivity, linearity of the gas sensitivity response curve, signal response time and recovery time. Generally speaking, the range of the gas sensor should be able to meet the measurement requirements of the maximum concentration of the characteristic gas leakage in a specific scenario, and the minimum concentration of the range should be small enough, that is, the maximum sensitivity of the gas sensor is large enough to detect the leakage of the characteristic gas in time at an early stage. The linearity of the gas sensor within the range should be good enough to reduce the measurement error. The response time and recovery time of the gas sensor should be short enough to realize real-time monitoring of the concentration change of the characteristic gas. Specifically, in this embodiment, (1) the range of the gas sensor includes at least 50 to 2000 ppm (2) the maximum sensitivity of the characteristic gas exceeds 15. (3) The linearity within the range is not less than 20%. (4) The signal response time T under the lowest concentration condition res No more than 10s. (5) Signal recovery time T rec No more than 100s.

[0058] The technicians set up a performance test experiment to test the 2 and SnO 2 The performance of the gas sensor with the composite material as the sensitive element was tested. The test results show that the gas sensitivity response curve of the gas sensor to EMC vapor with different concentration gradients is as follows: Figure 2 As shown. Figure 2 It can be found that this type of gas sensor can maintain good linearity in a wide range of 50-2000ppm, and also has good response characteristics under low concentration conditions, which is suitable for practical applications.

[0059] In addition, the performance test experiment also measured the response and recovery curve of the gas sensor to 500ppm EMC steam. The test results are as follows: Figure 3 As shown. Figure 3 It can be seen that the signal response time T of this type of gas sensor re s=3s, signal recovery time T rec =86S, both of which meet the preset indicator limits.

[0060] In summary, the gas sensor provided in this embodiment can meet the requirements in terms of measuring range, maximum sensitivity, linearity of gas sensitivity response curve, signal response time and recovery time.

[0061] Example 2

[0062] Based on the gas sensor with high precision and fast response characteristics to electrolyte vapor (DMC, DEC, EMC, etc.) provided in Example 1. This embodiment further provides a method for monitoring abnormalities of lithium batteries based on neuron detection, such as Figure 4 As shown in the figure, the solution deploys corresponding neural sensors at the air inlet, air outlet and detection sites corresponding to each battery module of the lithium battery cabinet to monitor the concentration of characteristic gases in real time. Then, according to the changes in the concentration of characteristic gases at different locations, it can quickly determine whether the lithium battery cabinet has safety risks.

[0063] Specifically, Figure 5 As shown, the abnormality monitoring method of lithium battery based on neuron detection provided in this embodiment includes three steps: abnormal signal capture, signal time domain analysis and environmental interference self-check, which specifically includes the following process:

[0064] 1. Abnormal signal capture

[0065] The real-time concentration Q of the characteristic gas at the air outlet, air inlet and detection points corresponding to each battery module of the lithium battery cabinet is collected in real time through the neural sensor, and the real-time concentration change rate V is calculated.

[0066] Among them, the calculation formula of the concentration change rate V of the characteristic gas is as follows:

[0067]

[0068] When the real-time concentration at any detection site reaches the preset safety threshold Qmax, an over-limit signal is generated and the over-limit time t0 is recorded.

[0069] Determine whether the duration ΔT and concentration change rate of the over-limit signal at each detection site have reached their respective preset safety thresholds: if yes, an abnormal state signal is generated; the abnormal state signal contains the state flag Abn and the coding information of the detection site. The generation formula of the state flag Abn of the abnormal state signal is as follows:

[0070]

[0071] In the above formula, V max Indicates the preset safety threshold of the characteristic gas concentration change rate; T max Indicates the preset safety threshold of the exceeding time limit.

[0072] The calculation formula of the duration ΔT of the over-limit signal is as follows:

[0073] ΔT=t1-t0

[0074] In the above formula, t1 represents the current time; t0 represents the time when the real-time concentration of the characteristic gas at the current detection point exceeds the safety threshold.

[0075] In the abnormal monitoring scheme of this embodiment, in order to reduce the false alarm rate, when the real-time concentration of the characteristic gas at a battery module in the battery cabinet exceeds the safety threshold, this embodiment will not immediately alarm, but will only generate a corresponding abnormal status signal when both the duration of the over-limit signal and the concentration change rate at that location exceed the preset safety threshold.

[0076] In the scheme of this real-time example, if any detection site generates an abnormal state signal, it means that the characteristic gas concentration at that location not only exceeds the safety threshold, but the concentration is actually still in an increasing state and has not dissipated for a long time. It can be seen that the conditions for triggering abnormal state signals designed in this embodiment are more scientific than the scheme using a single characteristic gas concentration threshold as a criterion in the traditional detection scheme, and can eliminate alarm phenomena caused by sensor abnormalities and accidental factors. In addition, on the basis of ensuring efficiency, the accuracy of the captured abnormal state signals is greatly improved and the false alarm rate is reduced.

[0077] 2. Signal time domain analysis

[0078] When an abnormal status signal is generated at any battery module, it is determined whether there is an abnormal status signal in the historical data at the air inlet before the preset delay period: if so, the abnormality investigation state is entered; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

[0079] In the scheme of this embodiment, it is assumed that when the fan speeds of the air inlet and the air outlet in the lithium battery cabinet are fixed, the path and time for the gas to reach each battery module from the air inlet are determined, and the time for any characteristic gas to travel from the air inlet to the detection site of each battery module is positively correlated with the length of the air duct from the detection site at each battery module to the air inlet. This embodiment defines the conveying time of the airflow in the air duct as a "preset delay period". At this time, if no abnormality occurs to the battery modules in the lithium battery cabinet, but there is electrolyte vapor in the external environment of the lithium battery cabinet, this will cause abnormal status signals to be generated successively at the air outlet and each battery module. At the same time, the time (or over-limit time) when the abnormal status signal is generated at each detection site should also meet the preset delay period.

[0080] On this basis, this embodiment first uses simulation or actual testing to determine the corresponding preset delay period at each detection site, and then when a battery module generates an abnormal state signal, the historical data of different detection sites are analyzed in the time domain in combination with the preset delay period. It should be emphasized that: the preset delay period can increase a certain percentage of margin on the basis of the measured value during actual application. For example, when it is determined through testing that the signal delay of a battery module A and the same characteristic gas sample measured at the air inlet is 1.0s, then in actual application, a 20% margin can be added to the signal delay, and the preset delay period is set to 1.2s. In this way, the query range of historical data at the air inlet can be expanded, thereby enhancing the ability of the present embodiment to distinguish abnormalities caused by environmental factors and eliminating random errors.

[0081] When an abnormal status signal is generated at any detection point in the lithium battery pack, first determine whether the outlet air has also generated an abnormal status signal before the corresponding preset delay period; if so, it means that the two abnormal status signals may be related, and further abnormality investigation is required in the subsequent process. If not, it means that the abnormal status signal at the battery module is caused by the leakage of its own electrolyte, not from the external environment. At this time, an alarm and disposal should be issued immediately.

[0082] In addition, during the signal time domain analysis stage, when abnormal status signals are present at the detection points at multiple battery modules and the air inlet in the lithium battery cabinet, it is also possible to determine whether the time difference between the over-limit moment at each battery module and the over-limit moment at the air inlet is positively correlated with the length of the air duct from each air inlet (i.e., consistent with the preset delay period). If so, the abnormality troubleshooting state is entered; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

[0083] When there is a safety risk of electrolyte leakage, this embodiment can also locate the abnormal battery module according to the coded information in the abnormal status signal, and then issue a risk warning signal containing location information so that staff can quickly reach the risk site for inspection and disposal.

[0084] 3. Environmental interference self-check

[0085] In the previous stage, when it is determined that the abnormal state signals of different monitoring sites are correlated in the time domain, it means that this may be an abnormality caused by environmental factors, but there is also the possibility that electrolyte leakage occurs at the corresponding time at different detection sites, thereby generating abnormal state signals. Therefore, the third stage of environmental interference self-checking is to further distinguish whether the consecutive alarm phenomena at different sites are environmental interference.

[0086] The abnormality troubleshooting process in this embodiment is as follows:

[0087] In the abnormal troubleshooting state, first close the air inlet and the air inlet fan, and execute an exhaust cycle of the preset duration.

[0088] Then determine whether the real-time concentration Q and concentration change rate V of the characteristic gas at the battery modules and air outlets with abnormal status signals during the exhaust cycle have dropped below the safety threshold: if so, it is determined that the abnormal status signal at the battery module is caused by an abnormal external environment; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

[0089] The method of abnormality troubleshooting in this embodiment is: first block the passage of gas from the external environment into the lithium battery cabinet, then exhaust the gas in the lithium battery cabinet, and finally measure whether the concentration of characteristic gas in the lithium battery cabinet gradually decreases to a normal value. If the abnormal status signals of all battery modules in the lithium battery cabinet can be completely removed after the above measures, it means that the high concentration of characteristic gas in the lithium battery cabinet is indeed flowing in from the external environment, and there is no need to issue a risk warning at this time.

[0090] In the environmental interference self-check stage, when the real-time concentration Q of the characteristic gas detected at any detection point is lower than the preset safety threshold Q after abnormality investigation and processing, max When the abnormal status signal of the current detection site needs to be updated, the status flag Abn is set from "1" to "0".

[0091] If the concentration of the characteristic gas still cannot be "cleared" after the above measures are taken, it means that the characteristic gas is actually caused by the electrolyte leakage of the battery module in the lithium battery cabinet, and the electrolyte leakage is still continuing. At this time, a risk warning must be issued immediately. After completing the abnormality investigation, when it is determined that there is a safety risk of electrolyte leakage, the abnormal battery module is located according to the coded information in the abnormal status signal, and then a risk warning signal containing the location information is issued.

[0092] In the solution provided in this embodiment, Figure 6 As shown, the method for locating the battery module with risks is as follows:

[0093] A unique coding information is assigned to the gas sensor installed at each detection site; and then the coding information and the state flag Abn of each detection site are sent out in the generated abnormal state signal.

[0094] When it is determined later that the battery modules at certain detection locations do have electrolyte leakage, the coding information in the abnormal state signal is first extracted; then the preset coding rule table is queried based on the coding information in the abnormal state signal, and the location information of the abnormal battery module is decoded to obtain.

[0095] In this embodiment, the coding rule table is a comparison table pre-established according to different gas sensor installation locations, which contains a one-to-one mapping relationship between coding information and location information. Figure 4 In the solution of single-layer symmetrical deployment of lithium battery modules, the coding information of pack1 can be recorded as "R01", representing the first battery module on the right close to the air inlet. The coding information of pack22 can be recorded as "L11", representing the 11th battery pack on the left away from the air inlet. Of course, in a lithium battery cabinet with a more complex spatial layout, other easily distinguishable coding rules can also be designed to encode the spatial coordinates of each battery module and generate corresponding coding information.

[0096] To sum up, the present embodiment provides a lithium battery abnormality monitoring method based on neuron detection, which can realize high-sensitivity real-time detection of electrolyte vapor in the lithium battery cabinet, and combine the "neuron detection technology" to realize spatiotemporal analysis of abnormal signals, eliminate pseudo-abnormal signals, reduce false alarm rate; and complete the rapid positioning of abnormal points.

[0097] Example 3

[0098] Based on the abnormality monitoring method of lithium batteries based on neuron detection designed in Example 2, this embodiment further provides an abnormality monitoring system for lithium battery cabinets based on a neuron sensor array. The monitoring system is a complete set of detection equipment that works based on the principle of the scheme in Example 2.

[0099] Specifically, Figure 7 As shown, the abnormality monitoring system of the lithium battery cabinet based on the neuron sensor array in this embodiment includes a neuron sensor array and a centralized analyzer. Among them, the neuron sensor array is composed of a plurality of gas sensors as in Example 1, and each gas sensor constitutes a neuron detector. According to the different installation positions of the gas sensors, the neuron detectors can be divided into two categories: Pack gas detectors and reference gas detectors. The Pack gas detector is installed at each battery module in the lithium battery cabinet to detect the real-time concentration of the characteristic gas at the battery module. The reference gas detector is installed at the air inlet and outlet of the battery cabinet to detect the real-time concentration of the characteristic gas at the air inlet and outlet of the lithium battery cabinet.

[0100] The centralized analyzer is electrically connected to each Pack gas detector and reference gas detector in the neuron sensor array. The real-time characteristic gas concentration detected by each gas sensor is obtained in real time, and then the abnormal monitoring method of lithium battery based on neuron detection in Example 1 is used to process the collected large amount of sample data, and then the real-time concentration of the characteristic gas at each detection point is used to determine whether the lithium battery cabinet has a safety risk of electrolyte leakage. Figure 8 This is a gas-sensitive response curve of a neuron sensor array containing 10 gas sensors when the concentration of characteristic gas increases.

[0101] In this embodiment, the centralized analyzer includes a data processing unit, a positioning unit and a communication unit. The data processing and analysis strategy of the data processing unit is as follows:

[0102] (1) Real-time collection of the real-time concentration Q detected by each gas sensor i , and use the following formula to calculate the real-time concentration change rate V i :

[0103]

[0104] (2) Determine the detection value Q of each gas sensor i When the safety threshold Qmax is reached, an over-limit signal is generated and the over-limit time t0 is recorded.

[0105] (3) Determine whether the duration ΔT and concentration change rate V of the over-limit signal of the gas sensor that has reached the safety threshold are both over the upper limit. If so, generate an abnormal state signal and record the coded information of the gas sensor where the abnormality occurs;

[0106] (4) Determine whether the gas sensor generating the abnormal state signal is a Pack gas detector. If so, query the preset delay period T of the Pack gas detector.

[0107] (5) Check whether there is an abnormal state signal at the air inlet in all time periods from the current moment to the preset delay period T: If there is, it is determined that the abnormality of the Pack gas detector may be caused by environmental factors, and the next step is entered; if not, it is determined that the electrolyte leakage has occurred at the current Pack gas detector, and there is a high safety risk.

[0108] (6) When it is determined that the abnormality of the Pack gas detector may be caused by environmental factors, close the air inlet and the air inlet fan, and execute an exhaust cycle of a preset duration;

[0109] (7) Obtaining the real-time concentration Q and concentration change rate V detected by each Pack gas detector and reference gas detector with abnormal state signals during the exhaust cycle;

[0110] (8) Determine whether the real-time concentration Q and concentration change rate V of the gas sensor with abnormal status signal have dropped below the safety threshold before the end of the exhaust cycle: if so, it is determined that the abnormal status signal at the battery module is caused by an abnormal external environment; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

[0111] The positioning unit is used to extract the corresponding coding information from the abnormal state signal generated by the Pack gas detector when the data processing unit determines that any battery module has a safety risk; then query a preset "code-position" comparison table to obtain the actual location of the battery module with a safety risk.

[0112] The communication unit is used to send the warning information representing the safety risk generated by the data processing unit and the location information of the battery module queried by the positioning unit to the control center or administrator of the lithium battery cabinet.

[0113] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring abnormalities of lithium batteries based on neuron detection, which detects the concentration of characteristic gases at different detection sites in the lithium battery cabinet in real time to determine whether there is a safety risk in the lithium battery cabinet; It is characterized in that The abnormality monitoring method comprises the following process:

1. Abnormal signal capture The real-time concentration Q of characteristic gases at the air outlet, air inlet and detection points corresponding to each battery module of the lithium battery cabinet is collected in real time through the neuron sensor, and the real-time concentration change rate V is calculated; The real-time concentration at any detection site reaches the preset safety threshold Q max When , an over-limit signal is generated and the over-limit time t0 is recorded; Determine whether the duration ΔT and concentration change rate of the over-limit signal at each detection site have reached their respective preset safety thresholds: if yes, generate an abnormal state signal, which includes a state flag Abn and coding information of the detection site; 2. Signal time domain analysis When an abnormal status signal is generated at any battery module, determine whether there is an abnormal status signal in the historical data before the preset delay period at the air inlet: if yes, enter the abnormality troubleshooting state; otherwise, determine that there is a safety risk of electrolyte leakage in the lithium battery cabinet; When there is a safety risk of electrolyte leakage, the abnormal battery module is located according to the coded information in the abnormal status signal, and then a risk warning signal containing the location information is issued; 3. Environmental interference self-check In the abnormal troubleshooting state, close the air inlet and air inlet fan, and execute the exhaust cycle of the preset duration; Determine whether the real-time concentration Q and concentration change rate V of the characteristic gas at the battery module and the air outlet with abnormal status signals during the exhaust cycle have dropped below the safety threshold: if so, it is determined that the abnormal status signal at the battery module is caused by an abnormal external environment; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet; When there is a safety risk of electrolyte leakage, the abnormal battery module is located according to the coded information in the abnormal status signal, and then a risk warning signal containing the location information is issued.

2. The method for abnormality monitoring of lithium batteries based on neuron detection according to claim 1, Features: In the abnormal signal capture stage, the concentration change rate is calculated as follows:

3. The abnormality monitoring method of lithium battery based on neuron detection as claimed in claim 2, It is characterized in that The process of locating the electrolyte leakage point is as follows: in the abnormal signal capture stage, each detection site corresponds to a unique coding information; the generated abnormal status signal contains the status flag Abn and coding information of each detection site; During the signal time domain analysis and environmental interference self-check stage, the preset coding rule table is queried according to the coding information in the abnormal state signal, and the location information of the abnormal battery module is decoded; a one-to-one mapping relationship between the coding information and the location information is established in the coding rule table.

4. The abnormality monitoring method of lithium battery based on neuron detection as claimed in claim 3, Features: In the abnormal signal capture stage, the generation formula of the status flag Abn in the abnormal status signal is as follows: In the above formula, V max Indicates the preset safety threshold of the characteristic gas concentration change rate; T max Indicates the preset safety threshold of the exceeding time limit.

5. The abnormality monitoring method of lithium battery based on neuron detection as claimed in claim 1, Features: In the signal time domain analysis stage, the preset delay period refers to the fluid delivery time from the air outlet to each battery module under the current fan speed conditions of the air inlet and outlet; the preset delay period of the detection point at each battery module in the lithium battery cabinet is positively correlated with the length of the air duct from the air inlet.

6. The method for abnormality monitoring of lithium batteries based on neuron detection as claimed in claim 1, Features: In the signal time domain analysis stage, when there are abnormal status signals at the detection points at multiple battery modules and the air inlet, it is determined whether the time difference between the over-limit time at each battery module and the over-limit time at the air inlet is positively correlated with the length of the air duct from each to the air inlet. If so, the abnormal investigation state is entered; otherwise, it is determined that there is a safety risk of electrolyte leakage in the lithium battery cabinet.

7. The abnormality monitoring method of lithium battery based on neuron detection as claimed in claim 1, Features: In the environmental interference self-check stage, when the real-time concentration Q of the characteristic gas detected at any detection point is lower than the preset safety threshold Q max When the abnormal status signal of the current detection site is updated.

8. An abnormal monitoring system for lithium battery cabinets based on a neural sensor array, It is characterized in that It includes: A neuron sensor array comprising a plurality of Pack gas detectors and a plurality of reference gas detectors; The Pack gas detector is installed at each battery module in the lithium battery cabinet to detect the real-time concentration of the characteristic gas at the battery module; the reference gas detector is installed at the air inlet and outlet of the battery cabinet to detect the real-time concentration of the characteristic gas at the air inlet and outlet of the lithium battery cabinet; a centralized analyzer is electrically connected to the neuron sensor array; the centralized analyzer includes a data processing unit, a positioning unit and a communication unit; the data processing unit uses the abnormal monitoring method of lithium batteries based on neuron detection as described in any one of claims 1 to 7 to process the collected sensor detection data, and then judges whether the lithium battery cabinet has a safety risk of electrolyte leakage according to the real-time concentration of the characteristic gas at each detection site; the positioning unit is used to query a "code-position" comparison table when any battery module has a safety risk, and determine the spatial position of the battery module; the communication unit is used to send out warning information representing the safety risk and the location information of the battery module.

9. The abnormality monitoring system for lithium battery cabinet based on neuron sensor array as claimed in claim 8, Features: The Pack gas detector and the reference gas detector both use a gas sensor specifically used to detect the concentration of characteristic gas containing any one or more components of dimethyl carbonate, diethyl carbonate, and ethyl methyl carbonate; The gas sensor comprises a sensitive element and a signal processing circuit; the sensitive element is made of TiO 2 and SnO 2 The signal processing circuit is used to convert the response of the sensitive element into a corresponding electrical signal.

10. The abnormality monitoring system for lithium battery cabinet based on neuron sensor array as claimed in claim 9, Features: The performance indicators of the gas sensor performance include: (1) The maximum sensitivity of the characteristic gas exceeds 15; (2) The linearity within the measuring range is not less than 20%; (3) Signal response time T under the lowest concentration condition res No more than 10 seconds; (4) Signal recovery time T rec No more than 100s.

Images