Lithium battery security system applied to electric bicycle garage

Abstract

The application belongs to the technical field of garage security, and provides a lithium battery security system applied to an electric bicycle garage, which comprises: dividing the electric bicycle garage into garage grids, setting up cameras in the garage grids to collect monitoring images to determine whether smoke exists in the garage grids, determining the cause of the smoke and taking targeted fire extinguishing measures for the smoke area if it is determined that the smoke exists, monitoring the charging of the lithium battery of the electric bicycle if it is determined that the smoke does not exist, calculating a comprehensive risk value of lithium battery thermal runaway and dividing a three-level response mechanism, the three-level response mechanism comprising normal monitoring, deep risk assessment and emergency power-off alarm, predicting the risk of lithium battery thermal runaway if the deep risk assessment is triggered, determining whether the deep monitoring is triggered, marking the deep monitoring camera and adjusting the camera collection parameters of the deep monitoring camera to conduct deep monitoring on the electric bicycle if the deep monitoring is triggered, and determining the monitoring busy degree in the garage grid and marking the deep monitoring camera.

Description

Technical Field

[0001] This invention belongs to the field of garage security technology, specifically a lithium battery security system for electric bicycle garages. Background Technology

[0002] With the increasing popularity of electric bicycles, the safety of electric bicycle garages, which serve as centralized parking and charging locations, is becoming increasingly prominent. The core component of electric bicycles—lithium batteries—poses a risk of thermal runaway and fire during charging and use. Once a fire occurs, it can spread rapidly and easily cause serious casualties and property damage. Therefore, how to effectively monitor and prevent safety hazards related to lithium batteries in electric bicycle garages has become an urgent technical problem to be solved.

[0003] In existing electric bicycle garage security technologies, the smoke detection aspect has many shortcomings. Some systems use single-point smoke sensors, which can only detect localized areas and cannot cover the entire garage space, easily leading to blind spots. Relying on manual patrols is inefficient and lacks real-time performance, failing to detect early smoke hazards in a timely manner. Some garages have installed cameras to monitor the situation inside; however, these cameras mainly focus on recording video of the overall garage environment and lack the ability to accurately identify smoke. When smoke appears in the garage, existing technologies often lack the ability to accurately determine the cause of the smoke. Whether the fire is caused by lithium battery thermal runaway or general combustible combustion, using the same fire extinguishing method may not only lead to poor fire extinguishing effects but may also cause secondary damage to lithium batteries due to the incorrect use of fire extinguishing agents. For example, using ordinary water-based fire extinguishing agents to extinguish lithium battery fires may exacerbate the battery's thermal runaway reaction and expand the accident's harm.

[0004] In the field of lithium battery charging monitoring, traditional methods mostly monitor only single parameters such as voltage and current, which cannot comprehensively assess the risk of thermal runaway of lithium batteries. Although some systems have risk warning mechanisms, they lack a scientific risk assessment system and cannot dynamically adjust monitoring strategies according to the actual state of lithium batteries. For low- to medium-risk lithium batteries, over-monitoring will lead to waste of resources, while for high-risk lithium batteries, untimely monitoring may lead to accidents. In addition, in terms of monitoring resource allocation, existing technologies do not fully consider the actual usage of monitoring equipment in garages, making it difficult to achieve accurate and focused monitoring of high-risk lithium batteries.

[0005] To address the above problems, this invention proposes a lithium battery security system for electric bicycle garages. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve the technical problem is: a lithium battery security system for electric bicycle garages, comprising:

[0008] Global monitoring module: Divides the electric bicycle garage into garage grids, and sets up cameras within the garage grids to collect monitoring images to determine whether there is smoke within the garage grids;

[0009] Specifically, for the cameras within the garage grid, the camera acquisition parameters are set and initialized. Based on the camera acquisition parameters, corresponding monitoring images are acquired in real time. The monitoring images are then analyzed using the HSV color space to filter out suspected smoke areas.

[0010] Edge detection is performed on suspected smoke areas within the monitoring image to determine edge points. A frame segment of the monitoring image is extracted, including several frames of the monitoring image preceding the current image. The characteristics of the edge points of the corresponding suspected smoke areas within the frame segment are analyzed to filter out smoke areas.

[0011] If a smoke area is present in any monitoring image within the garage grid at the current time, it is determined that smoke exists within the garage grid; otherwise, it is determined that smoke does not exist within the garage grid.

[0012] For the processing module: If smoke is present, determine the cause of the smoke and take targeted fire extinguishing measures for the smoke-affected area;

[0013] Specifically, if it is determined that there is smoke in the garage grid, real-time feature extraction is performed on the smoke area, and the real-time smoke feature vector of the smoke area is collected.

[0014] A multi-dimensional lithium battery thermal runaway smoke feature library was established. Smoke feature vectors of several sets of smoke samples were collected through lithium battery thermal runaway simulation experiments. Clustering algorithm was used to classify the feature vectors, and data processing was performed on each type of feature vector to obtain several sets of standard feature vectors.

[0015] Calculate the Euclidean distance between the real-time smoke feature vector and the standard feature vector. If any Euclidean distance is less than or equal to the Euclidean distance threshold, the smoke in the smoke area is determined to be caused by lithium battery thermal runaway. Otherwise, the smoke in the smoke area is determined to be caused by general combustibles.

[0016] Using stereo vision from a binocular camera, the smoke area is visually located, the three-dimensional coordinates of the center point of the smoke area are calculated, and the cause of the smoke in the smoke area is obtained.

[0017] If the smoke is caused by common combustibles, the risk level of the smoke area is judged by the location of the center point of the smoke area and the size of the smoke area. If the risk level is high, the sprinkler heads are controlled to spray water to extinguish the fire in the smoke area.

[0018] If the smoke is caused by thermal runaway of the lithium battery, the nearest neighbor algorithm is used to match the location of the lithium battery and calculate the three-dimensional coordinates of the center point of the lithium battery. The fire extinguishing agent nozzle is then controlled to spray a new type of lithium battery fire extinguishing agent at the coordinates of the center point of the lithium battery.

[0019] Risk prediction module: If there is no smoke, the charging monitoring of the lithium battery of the electric bicycle is carried out, the comprehensive risk value of lithium battery thermal runaway is calculated and divided into three-level response mechanisms. The three-level response mechanisms include normal monitoring, in-depth risk assessment and emergency power failure alarm. If in-depth risk assessment is triggered, the comprehensive risk value of lithium battery thermal runaway is predicted, and the in-depth monitoring is determined based on the prediction results.

[0020] Specifically, if it is determined that there is no smoke in the garage grid, the charging voltage and charging current of each charging port in the garage grid are collected in real time, and the ambient temperature is collected in real time. The voltage deviation, current deviation and temperature deviation are calculated in combination with key safety thresholds.

[0021] Set a historical risk accumulation factor and assign it an initial value of 0, and set a risk reduction period;

[0022] If the charging voltage, charging current, and ambient temperature exceed the critical safety threshold at the current time, the historical risk accumulation factor will increment once; if the collected charging voltage, charging current, and ambient temperature do not exceed the critical safety threshold within the continuous risk reduction period, the historical risk accumulation factor will decrement once.

[0023] The historical risk accumulation factors at each time point during the charging period are integrated into a risk sequence. Based on the risk sequence, the historical risk accumulation factors at the current time are normalized to obtain the normalized risk accumulation factor.

[0024] The analytic hierarchy process (AHP) was used to calculate the weights of voltage deviation, current deviation, temperature deviation, and normalized risk accumulation factor. The data was then processed in conjunction with the weights to obtain the comprehensive risk value of the lithium battery of the electric bicycle that was being charged.

[0025] If the overall risk value is less than the medium-level risk range, normal monitoring is triggered; if it is greater than the medium-level risk range, an emergency power outage alarm is triggered; if it falls within the medium-level risk range, an in-depth risk assessment is triggered.

[0026] If a deep risk assessment is triggered, the comprehensive risk value of the electric bicycle lithium battery during the charging period is obtained and integrated into a comprehensive risk value sequence according to the time series. Based on the comprehensive risk value sequence, the ARIMA model is used to predict the comprehensive risk value during the risk prediction period.

[0027] If the predicted comprehensive risk value exceeds the medium-level risk range during the risk prediction period, in-depth monitoring of the lithium battery of electric bicycles will be triggered.

[0028] Deep monitoring module: If deep monitoring is triggered, determine the monitoring busy level within the garage grid and mark the deep monitoring camera, then adjust the camera acquisition parameters of the deep monitoring camera to perform deep monitoring of electric bicycles;

[0029] Specifically, if the camera acquisition parameters of all cameras in the garage grid are the same as the initial values, the cameras are marked as global cameras. If the number of global cameras in the garage grid is greater than 1, the three-dimensional coordinates of the center point of the electric bicycle are calculated for visual positioning.

[0030] The global camera with the smallest distance from the center point of the electric bicycle is marked as the depth monitoring camera. The camera acquisition parameters of the depth monitoring camera are adjusted to focus on the electric bicycle for monitoring.

[0031] The beneficial effects of this invention are as follows:

[0032] 1. This invention divides electric bicycle garages into grids and sets up cameras to accurately collect monitoring images, which can quickly determine whether there is smoke in the grid. Once smoke is detected, it can not only determine the cause and locate the area, but also quickly take targeted fire extinguishing measures to effectively curb the spread of fire, greatly reduce the losses caused by fire, and protect the safety of people and property in the garage. Compared with traditional security methods, the response to smoke is more timely and the handling is more accurate, building a solid defense for garage safety.

[0033] 2. This invention monitors lithium batteries during charging in smoke-free conditions, calculates the comprehensive risk value of thermal runaway, and divides it into three levels of response mechanisms. When a deep risk assessment is triggered, the risk value can be predicted, and the decision on whether to conduct deep monitoring is made based on the prediction results. At the same time, the deep monitoring cameras are marked according to the monitoring busy level and the parameters are adjusted to achieve more detailed monitoring of electric bicycles. This comprehensive and intelligent management greatly improves the security level of lithium batteries in garages and ensures the safe and stable operation of garages. Attached Figure Description

[0034] The invention will now be further described with reference to the accompanying drawings.

[0035] Figure 1 This is a schematic diagram of a lithium battery security system module for electric bicycle garages, as described in an embodiment of the present invention.

[0036] Figure 2 This is a flowchart illustrating the specific steps involved in triggering deep monitoring in a lithium battery security system for electric bicycle garages, as described in an embodiment of the present invention. Detailed Implementation

[0037] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0038] Example 1

[0039] Please see Figure 1 As shown in the embodiment of the present invention, a lithium battery security system for electric bicycle garages includes the following modules:

[0040] Global monitoring module: Divides the electric bicycle garage into garage grids, sets up cameras within the garage grids to collect monitoring images, and determines whether there is smoke within the garage grids based on the monitoring images;

[0041] In electric bicycle garages, a grid-based deployment strategy is adopted based on the garage area and structural layout. The garage is divided into several garage grids in the horizontal direction, and several cameras are set in each garage grid and installed on the top of the corresponding garage grid to ensure that there are no blind spots in the monitoring inside the garage.

[0042] It should be noted that if the electric bicycle garage is a small garage, the number of garage grids in the garage can be 1. Each garage grid is equipped with several cameras, and the monitoring area of ​​each camera can cover the entire garage grid. The purpose of setting up several cameras is to improve the monitoring accuracy.

[0043] Configure camera acquisition parameters, including resolution, frame rate, exposure time, shooting angle, and magnification. Perform the same initialization settings for all cameras in the garage. Connect the cameras to the garage's internal network through the network configuration module, and configure network parameters such as IP address, subnet mask, and gateway to ensure stable communication between the cameras and the server. Use an NTP (Network Time Protocol) server for time synchronization to ensure time consistency of images acquired by all cameras in the garage, facilitating subsequent event tracing and analysis.

[0044] For any garage grid, the camera inside the garage grid collects the corresponding monitoring image in real time according to the camera acquisition parameters. The monitoring image is converted into HSV (hue, saturation, brightness) color space. In the HSV color space, the saturation range and brightness range are set. Adjacent pixels whose saturation and brightness both fall within the corresponding saturation range and brightness range are grouped into a pixel group. The area where the pixel group contains the number of pixels that meets the recognition conditions is marked as a suspected smoke area.

[0045] It should be noted that in the HSV color space, smoke usually exhibits characteristics of low saturation and medium brightness. Therefore, the HSV color space can be used to initially screen suspected smoke areas.

[0046] Edge detection is performed on suspected smoke areas within the monitoring image using the Canny edge detection algorithm. The monitoring image is smoothed by Gaussian filtering, and the gradient magnitude and direction of the suspected smoke areas within the monitoring image are calculated. Non-maximum suppression and double threshold detection are applied to determine the edge points of the suspected smoke areas. The connectivity between edge points is calculated, and the continuous length between adjacent edge points is statistically analyzed based on the connectivity between edge points and marked as continuous length. The mean of all continuous lengths in the suspected smoke area is calculated to obtain the average continuous length.

[0047] Based on any suspected smoke area in the current monitoring image, obtain the suspected smoke area in the previous frame monitoring image that overlaps with the suspected smoke area, compare the average continuous length of the two suspected smoke areas and take the absolute difference as the continuous length change value. If there is no suspected smoke area in the previous frame monitoring image that overlaps with the suspected smoke area, assign the continuous length change value as the average continuous length of the suspected smoke area.

[0048] Extract a frame segment of a monitoring image, including several frames of monitoring images preceding the current monitoring image. Take the average value of the continuous length variation of the corresponding suspected smoke area within the frame segment to obtain the average continuous length variation of the corresponding suspected smoke area in the current monitoring image. Set the average continuous length range and the average continuous length variation range. Mark the suspected smoke area where both the average continuous length and the average continuous length variation fall within the corresponding range as a smoke area.

[0049] If a smoke area is present in any monitoring image within the garage grid at the current time, it is determined that smoke exists within the garage grid; otherwise, it is determined that smoke does not exist within the garage grid.

[0050] It should be noted that the function of this module is to achieve automated and high-precision detection of smoke in the garage. Compared with manual inspection, it can detect potential smoke hazards earlier and more accurately, cover all areas of the garage, avoid monitoring blind spots, buy time for timely handling of fire hazards, and reduce the probability of fire.

[0051] For the processing module: if smoke is detected within the garage grid, determine the cause of the smoke, locate the smoke area, and take targeted fire extinguishing measures for the smoke area;

[0052] If smoke is detected in the garage grid, real-time feature extraction is performed on the smoke area based on the smoke area in the monitoring image. Smoke feature parameters of the smoke area are collected and integrated into a real-time smoke feature vector X. The smoke feature parameters include the average gray value of the smoke area, the temperature of the smoke area, and the carbon monoxide concentration of the smoke area. The average gray value of the smoke area is calculated by the gray value of each pixel in the smoke area in the monitoring image. The temperature of the smoke area is collected by an infrared temperature measurement array deployed on the garage ceiling. The carbon monoxide concentration of the smoke area is collected by an electrochemical sensor.

[0053] A multi-dimensional lithium battery thermal runaway smoke feature library was established. Smoke feature parameters of several sets of smoke samples were collected through lithium battery thermal runaway simulation experiments. These parameters were then integrated into smoke feature vectors. The K-means clustering algorithm was used to classify the feature vectors, and the mean of each class was calculated to obtain several sets of standard feature vectors. These standard feature vectors were then integrated to form a standard smoke feature set.

[0054] ;

[0055] in, This represents the nth set of standard eigenvectors. N represents the number of standard feature vectors in the standard smoke feature set;

[0056] Calculate the Euclidean distance between the real-time smoke feature vector X and the standard feature vector in the standard smoke feature set. If there exists any standard feature vector whose Euclidean distance to the real-time smoke feature vector X is less than or equal to the Euclidean distance threshold, then it is determined that the generation of the smoke region is caused by the thermal runaway of the lithium battery; otherwise, it is determined that the generation of the smoke region is caused by general combustibles.

[0057] Visual positioning of the smoke area is achieved using stereo vision from a binocular camera, and the three-dimensional coordinates of the center point of the smoke area are calculated.

[0058] For smoke areas caused by common combustibles, obtain the shortest distance between the center point of the smoke area and the parked electric bicycle, and calculate the number of pixels in the smoke area using the current monitoring image. If the shortest distance is less than the safe distance threshold and the number of pixels is greater than the number of pixels threshold, it is determined that the smoke area is at high risk, and the sprinkler head set on the garage ceiling closest to the center point of the smoke area is controlled to spray water to extinguish the fire in the smoke area.

[0059] For the smoke area caused by lithium battery thermal runaway, the nearest neighbor algorithm is used to match the most likely lithium battery location based on the calculated three-dimensional coordinates of the smoke area's center point. The three-dimensional coordinates of the lithium battery's center point are then calculated. The fire extinguishing agent nozzle, located on the garage ceiling closest to the smoke area's center point, is then controlled. The rotation angle of the fire extinguishing agent nozzle is calculated based on the lithium battery's center point's three-dimensional coordinates. A PID control algorithm is then used to precisely adjust the nozzle angle. After adjustment, the fire extinguishing agent nozzle is controlled to spray a new type of lithium battery fire extinguishing agent targeting the lithium battery's center point coordinates, preventing fires caused by lithium battery thermal runaway in the electric bicycle garage.

[0060] The preparation method of the novel lithium battery fire extinguishing agent is as follows: take 4-7 parts by weight of flame retardant, 10-14 parts by weight of adsorbent, 5-8 parts by weight of sodium bicarbonate, 18-21 parts by weight of foaming agent, 11-13 parts by weight of foam stabilizer, and 42-53 parts by weight of water. Mix the foaming agent, foam stabilizer, and water, heat and stir, filter, and obtain intermediate product A. Add adsorbent and flame retardant to intermediate product A, stir, cool, and obtain intermediate product B. Add sodium bicarbonate to intermediate product B, stir, and obtain the novel lithium battery fire extinguishing agent. The flame retardant is a mixture of (trimethylsilyl) phosphite and zinc borate, and the adsorbent is a mixture of aluminum hydroxide, activated carbon, and modified vermiculite. The modified vermiculite is prepared by wet milling of vermiculite, mixing with modifiers, dispersants, etc., and then treating with a combination of ultrasonic and ultra-high frequency ultraviolet light.

[0061] It should be noted that the purpose of this module is to accurately distinguish the source of smoke and avoid ineffective handling due to misjudgment. For fires caused by thermal runaway of lithium batteries, the new lithium battery fire extinguishing agent is used to extinguish the fire accurately, reduce damage to the lithium battery, and prevent the fire from spreading. For general combustible fires, conventional fire extinguishing methods are used to respond quickly, improve fire extinguishing efficiency, and ensure the safety of garages and vehicles.

[0062] Risk prediction module: If it is determined that there is no smoke in the garage grid, the charging monitoring of the lithium battery of the electric bicycle that is charging is carried out, the comprehensive risk value of lithium battery thermal runaway is calculated and a three-level response mechanism is divided. The three-level response mechanism includes normal monitoring, in-depth risk assessment and emergency power failure alarm. If in-depth risk assessment is triggered, the comprehensive risk value of lithium battery thermal runaway is predicted, and the in-depth monitoring is determined based on the prediction result.

[0063] like Figure 2 As shown, the specific steps for triggering deep monitoring are as follows;

[0064] If it is determined that there is no smoke in the garage grid, the charging process of the lithium battery of the electric bicycle in the garage grid is analyzed through the charging port in the garage grid;

[0065] Specifically, each charging port in the garage grid is equipped with a high-precision voltage and current sensor to collect charging voltage data in real time. With charging current This forms a charging parameter sequence:

[0066] ;

[0067] Where i represents the charging port number and t represents the current time;

[0068] Based on the safety parameters of common electric bicycle lithium batteries, key safety thresholds are set, including the upper limit of the rated charging voltage. Upper limit of rated charging current and safe operating temperature range ;

[0069] Based on the actual collected charging voltage With charging current Calculate voltage deviation by combining key safety thresholds Deviation from current The formula is:

[0070] ;

[0071] ;

[0072] in, Indicates the rated voltage. Indicates the rated current;

[0073] The ambient temperature within the garage grid is monitored using a temperature sensor. Real-time data collection is performed, and temperature deviation is calculated based on key safety thresholds. The formula is:

[0074] ;

[0075] in, This indicates the preset optimal operating temperature of the lithium battery;

[0076] Set a historical risk accumulation factor and assign it an initial value of 0, and set a risk reduction period;

[0077] If the collected charging voltage, charging current and ambient temperature exceed the critical safety threshold, the historical risk accumulation factor will be incremented once.

[0078] If the collected charging voltage, charging current and ambient temperature do not exceed the critical safety threshold within a continuous period of risk reduction, the historical risk accumulation factor will undergo a self-decrease operation.

[0079] The time period from the start time of charging the lithium battery of the electric bicycle to the current time is marked as the charging period. The historical risk accumulation factor at each time within the charging period is integrated into a risk sequence. Based on the risk sequence, the historical risk accumulation factor at the current time is normalized to obtain the normalized historical risk accumulation factor, which is marked as the normalized risk accumulation factor.

[0080] Using the analytic hierarchy process (AHP), a judgment matrix is ​​constructed based on the calculated voltage deviation, current deviation, temperature deviation, and normalized risk accumulation factor. The corresponding weight vector is calculated based on the judgment matrix, and a consistency check is performed. Finally, the assigned weights of voltage deviation, current deviation, temperature deviation, and normalized risk accumulation factor are calculated. The weighted sum of voltage deviation, current deviation, temperature deviation, and normalized risk accumulation factor is combined with the assigned weights to obtain the comprehensive risk value of the lithium battery of the electric bicycle that is being charged at the charging port.

[0081] A three-tiered response mechanism is established based on the comprehensive risk value, with a medium-level risk range defined.

[0082] If the overall risk value is less than the medium-level risk range, the risk of thermal runaway of the electric bicycle lithium battery is judged to be low. If the risk of thermal runaway of the electric bicycle lithium battery can be judged to be low for a continuous period of risk reduction, and deep monitoring of the electric bicycle lithium battery has been triggered, then deep monitoring is exited.

[0083] If the overall risk value is greater than the medium-level risk range, it is determined that the risk of thermal runaway of the electric bicycle lithium battery is high. Immediately stop charging the electric bicycle lithium battery and generate an alarm signal to send to the electric bicycle owner's mobile phone to remind the owner to repair the electric bicycle lithium battery. If deep monitoring of the electric bicycle lithium battery has been triggered, exit deep monitoring.

[0084] If the overall risk value falls within the medium-level risk range, it is determined that the electric bicycle has a medium-level risk of lithium battery thermal runaway. The electric bicycle lithium battery will continue to be charged, and a deep risk assessment will be triggered for the electric bicycle lithium battery.

[0085] For any electric bicycle lithium battery, if a deep risk assessment is triggered, the remaining charging time of the electric bicycle lithium battery is calculated in real time.

[0086] Specifically, charging the electric bicycle within the remaining demand time is considered as constant current charging, and the formula for calculating the remaining charging time is:

[0087] ;

[0088] in, Indicates the remaining charging time. This indicates the nominal capacity of the lithium battery in the electric bicycle. This represents the charging current obtained in real time from the charging port i of the electric bicycle's lithium battery at the current time t. This represents the charging efficiency factor. The formula for calculating the current charge level of an electric bicycle is as follows:

[0089] ;

[0090] in, This indicates the start of the charging period, and t represents the current time. Indicates time;

[0091] The time period starting from the current time and with the remaining charging time as the duration will be marked as the risk prediction period.

[0092] It should be noted that the remaining charging time will vary over time, so the duration of the risk prediction period may not be the same at different times. The purpose of calculating the remaining charging time is to dynamically determine the risk prediction period and synchronize the monitoring strategy with the charging process.

[0093] The comprehensive risk value of the lithium battery of the electric bicycle during the charging period is obtained and integrated into a comprehensive risk value sequence according to the time series. Based on the comprehensive risk value sequence, a method is adopted. The model predicts the comprehensive risk value during the risk prediction period, where p is the autoregression order, d is the difference order, and q is the moving average order.

[0094] Specifically, the stationarity of the comprehensive risk value series is tested. If it is not stationary, the series is differencing to make it stationary. The optimal parameters p, d, and q of the model are determined by minimizing the Bayesian information criterion. The specific prediction formula is as follows:

[0095] ;

[0096] in, This represents the predicted composite risk value at time t+k. and These represent the autoregressive coefficient and the moving average coefficient, respectively. This represents the residual over time t+k−j, where t+k represents the time period within the risk prediction period.

[0097] If the predicted comprehensive risk value exceeds the medium-level risk range during the risk prediction period, in-depth monitoring of the lithium battery of electric bicycles will be triggered.

[0098] It should be noted that the function of this module is to realize dynamic risk assessment of the lithium battery charging process, predict the risk of thermal runaway in advance, and rationally allocate monitoring resources through a graded response mechanism. For low-risk lithium batteries, unnecessary resource occupation is reduced, and for high-risk lithium batteries, the power is cut off in time and the vehicle owner is notified, effectively preventing lithium battery thermal runaway accidents. At the same time, the in-depth assessment of medium-risk lithium batteries is triggered to ensure that the risk is under control.

[0099] Deep monitoring module: If deep monitoring is triggered, determine the monitoring busy level within the garage grid and mark the deep monitoring camera, then adjust the camera acquisition parameters of the deep monitoring camera to perform deep monitoring of electric bicycles;

[0100] If deep monitoring of the electric bicycle's lithium battery is triggered, determine the monitoring busy level within the garage grid.

[0101] Specifically, obtain the acquisition parameters of all cameras within the garage grid. For any camera within the garage grid, if the acquisition parameters of the camera are all the same as the initial value setting, mark the camera as a global camera.

[0102] Get the number of global cameras in the garage grid. If the number of global cameras is 1, determine that the monitoring is busy and generate a monitoring busy signal.

[0103] If the number of cameras in the entire system is greater than 1, the monitoring busy level is considered low.

[0104] It should be noted that, in order to ensure global monitoring coverage of the garage grid, there is no situation where the number of global cameras is less than 1;

[0105] If the monitoring workload is low, stereo vision from a binocular camera can be used to visually locate the center point of the electric bicycle and calculate its three-dimensional coordinates. ;

[0106] Obtain the 3D coordinates of all global cameras , where j represents the number of the global camera;

[0107] Calculate the distance between the 3D coordinates of all global cameras and the 3D coordinates of the center point of the electric bicycle, and mark the global camera with the smallest calculated distance as the depth monitoring camera of the electric bicycle.

[0108] For depth surveillance cameras, adjust the camera acquisition parameters of the depth surveillance cameras;

[0109] Specifically, the shooting angle in the camera acquisition parameters is adjusted based on the three-dimensional coordinates of the electric bicycle's center point and the three-dimensional coordinates of the depth monitoring camera, and the horizontal deflection angle is calculated. With pitch angle The formula is:

[0110] ;

[0111] ;

[0112] Based on the calculated horizontal deflection angle With pitch angle The shooting angle of the depth monitoring camera is adjusted by using a PID algorithm so that the center point of the electric bicycle falls on the center point of the monitoring image captured by the depth monitoring camera. The magnification in the camera acquisition parameters is also adjusted to focus on the electric bicycle, so that the depth monitoring camera can monitor the lithium battery of the electric bicycle.

[0113] If you want to exit the deep monitoring of the electric bicycle's lithium battery, initialize the camera acquisition parameters of the deep monitoring camera.

[0114] It should be noted that the function of this module is to rationally allocate camera resources based on the actual monitoring resources, so as to ensure key and accurate monitoring of medium and high risk lithium batteries. While ensuring the monitoring effect, it avoids excessive occupation of monitoring resources and improves resource utilization efficiency. By dynamically adjusting the camera parameters, it can clearly capture changes in the state of lithium batteries, providing clear image evidence for timely detection and handling of potential risks, and further enhancing the reliability of the lithium battery security system.

[0115] The technical solution of this invention is as follows: The electric bicycle garage is divided into garage grids. Cameras are installed within the garage grids to collect monitoring images. Based on the monitoring images, it is determined whether smoke exists within the garage grid. If smoke is detected, the cause of the smoke is determined, the smoke area is located, and targeted fire extinguishing measures are taken. If no smoke is detected, the charging of the lithium battery of the electric bicycle is monitored. The comprehensive risk value of lithium battery thermal runaway is calculated, and a three-level response mechanism is established. The three-level response mechanism includes normal monitoring, in-depth risk assessment, and emergency power outage alarm. If in-depth risk assessment is triggered, the comprehensive risk value of lithium battery thermal runaway is predicted. Based on the prediction result, it is determined whether in-depth monitoring is triggered. If in-depth monitoring is triggered, the monitoring busy level within the garage grid is determined, and the in-depth monitoring cameras are marked. The camera acquisition parameters of the in-depth monitoring cameras are adjusted to perform in-depth monitoring of the electric bicycle.

[0116] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A lithium battery security system for electric bicycle garages, characterized in that: include: Global monitoring module: Divides the electric bicycle garage into garage grids, and sets up cameras within the garage grids to collect monitoring images to determine whether there is smoke within the garage grids; For the processing module: If smoke is present, determine the cause of the smoke and take targeted fire extinguishing measures for the smoke-affected area; The method for obtaining the cause of the smoke is as follows: If it is determined that there is smoke in the garage grid, real-time feature extraction is performed on the smoke area, and the real-time smoke feature vector of the smoke area is collected. A multi-dimensional lithium battery thermal runaway smoke feature library was established. Smoke feature vectors of several sets of smoke samples were collected through lithium battery thermal runaway simulation experiments. Clustering algorithm was used to classify the feature vectors, and data processing was performed on each type of feature vector to obtain several sets of standard feature vectors. Calculate the Euclidean distance between the real-time smoke feature vector and the standard feature vector. If any Euclidean distance is less than or equal to the Euclidean distance threshold, the smoke in the smoke area is determined to be caused by lithium battery thermal runaway. Otherwise, the smoke in the smoke area is determined to be caused by general combustibles. Risk prediction module: If there is no smoke, the charging monitoring of the lithium battery of the electric bicycle is carried out, the comprehensive risk value of lithium battery thermal runaway is calculated and divided into three-level response mechanisms. The three-level response mechanisms include normal monitoring, in-depth risk assessment and emergency power failure alarm. If in-depth risk assessment is triggered, the comprehensive risk value of lithium battery thermal runaway is predicted, and the in-depth monitoring is determined based on the prediction results. The method for obtaining the comprehensive risk value is as follows: If it is determined that there is no smoke in the garage grid, the charging voltage and charging current of each charging port in the garage grid are collected in real time, and the ambient temperature is collected in real time. The voltage deviation, current deviation and temperature deviation are calculated in combination with the key safety thresholds. The normalized risk accumulation factor is obtained, and the weights of voltage deviation, current deviation, temperature deviation and normalized risk accumulation factor are calculated using the analytic hierarchy process. The data is then processed in combination with the weights to obtain the comprehensive risk value of the lithium battery of the electric bicycle that is being charged at the charging port. The normalized risk cumulative factor is obtained as follows: Set a historical risk accumulation factor and assign it an initial value of 0, and set a risk reduction period; If the charging voltage, charging current, and ambient temperature exceed the critical safety threshold at the current time, the historical risk accumulation factor will increment once; if the collected charging voltage, charging current, and ambient temperature do not exceed the critical safety threshold within the continuous risk reduction period, the historical risk accumulation factor will decrement once. The historical risk accumulation factors at each time point during the charging period are integrated into a risk sequence. Based on the risk sequence, the historical risk accumulation factors at the current time are normalized to obtain the normalized risk accumulation factor. Deep monitoring module: If deep monitoring is triggered, the module determines the monitoring busy level within the garage grid and marks the deep monitoring cameras. It then adjusts the camera acquisition parameters of the deep monitoring cameras to perform deep monitoring of electric bicycles.

2. The lithium battery security system for electric bicycle garages according to claim 1, characterized in that: The method for determining whether smoke exists within the garage grid is as follows: For the cameras within the garage grid, set the camera acquisition parameters and perform initialization settings. Acquire corresponding monitoring images in real time according to the camera acquisition parameters, perform HSV color space analysis on the monitoring images, and filter out suspected smoke areas. Edge detection is performed on suspected smoke areas within the monitoring image to determine edge points. A frame segment of the monitoring image is extracted, including several frames of the monitoring image preceding the current image. The characteristics of the edge points of the corresponding suspected smoke areas within the frame segment are analyzed to filter out smoke areas. If a smoke area is present in any monitoring image within the garage grid at the current time, it is determined that smoke exists within the garage grid; otherwise, it is determined that smoke does not exist within the garage grid.

3. The lithium battery security system for electric bicycle garages according to claim 1, characterized in that: The targeted firefighting measures include: Using stereo vision from a binocular camera, the smoke area is visually located, the three-dimensional coordinates of the center point of the smoke area are calculated, and the cause of the smoke in the smoke area is obtained. If the smoke is caused by common combustibles, the risk level of the smoke area is judged by the location of the center point of the smoke area and the size of the smoke area. If the risk level is high, the sprinkler heads are controlled to spray water to extinguish the fire in the smoke area. If the smoke is caused by thermal runaway of the lithium battery, the nearest neighbor algorithm is used to match the location of the lithium battery and calculate the three-dimensional coordinates of the center point of the lithium battery. The fire extinguishing agent nozzle is then controlled to spray a new type of lithium battery fire extinguishing agent at the coordinates of the center point of the lithium battery.

4. A lithium battery security system for electric bicycle garages according to claim 1, characterized in that: The method for determining whether deep monitoring has been triggered is as follows: The comprehensive risk value of the lithium battery of electric bicycle during the charging period is obtained and integrated into a comprehensive risk value sequence according to the time series. Based on the comprehensive risk value sequence, the ARIMA model is used to predict the comprehensive risk value during the risk prediction period. If the predicted comprehensive risk value exceeds the medium-level risk range during the risk prediction period, in-depth monitoring of the lithium batteries of electric bicycles will be triggered.

5. A lithium battery security system for electric bicycle garages according to claim 4, characterized in that: The method for obtaining the risk prediction period is as follows: The charging of electric bicycles within the remaining demand time is considered as constant current charging. The remaining charging time of the electric bicycle's lithium battery is calculated in real time. Starting from the current time, a period of time with the same duration as the remaining charging time is extracted to obtain the risk prediction period.

6. A lithium battery security system for electric bicycle garages according to claim 1, characterized in that: The three-level response mechanism is divided as follows: Obtain the comprehensive risk value of the lithium battery of the electric bicycle. If the comprehensive risk value is less than the medium-level risk range, normal monitoring is triggered. If it is greater than the medium-level risk range, an emergency power-off alarm is triggered. If it falls within the medium-level risk range, a deep risk assessment is triggered.

7. A lithium battery security system for electric bicycle garages according to claim 1, characterized in that: The deep monitoring includes: If the camera acquisition parameters of all cameras in the garage grid are the same as the initial values, the cameras are marked as global cameras. If the number of global cameras in the garage grid is greater than 1, the three-dimensional coordinates of the center point of the electric bicycle are calculated for visual positioning. The global camera with the smallest distance from the center point of the electric bicycle is marked as the depth monitoring camera. The camera acquisition parameters of the depth monitoring camera are adjusted to focus on the electric bicycle for monitoring.

Images