Power distribution method and system based on multiple cameras

Abstract

The invention relates to the technical field of camera power distribution, in particular to a power distribution method and system based on multiple cameras, and the method comprises the following steps: S1, initializing a power distribution system; s2, dynamic power dispatching; s3, data synchronization and transmission; s4, fault detection and processing; s5, intelligent load prediction: predicting the power demand of each camera in a period of time in the future by analyzing historical power use data and a current working state; based on a prediction result, a power distribution strategy is adjusted in advance, and the sudden power shortage or overload condition is avoided; and the accuracy of load prediction is optimized in combination with external data of weather forecast. According to the method, historical power use data and a current working state are analyzed through intelligent load prediction, and power demand prediction is performed by adopting an improved time sequence analysis method and an auto-encoder model in combination with weather forecast data. And according to a prediction result, a power distribution strategy is adjusted in advance, power supply of the key camera is guaranteed preferentially, and the sudden power shortage or overload condition is avoided.

Description

Technical Field

[0001] The present invention relates to the field of camera power distribution technology, and in particular to a power distribution method and system based on multiple cameras. Background Art

[0002] In modern intelligent monitoring systems, multi-camera systems are widely used in transportation hubs, banks, airports, industrial parks and other fields for real-time monitoring and management of complex environments. However, the existing multi-camera-based power distribution methods have many shortcomings, such as low power efficiency, untimely fault detection, unstable data transmission, etc. These problems seriously affect the overall performance and reliability of the monitoring system.

[0003] In the existing technology, the power distribution of cameras usually adopts a fixed power distribution scheme, which cannot dynamically adjust the power distribution to cope with the changes in real-time demand. In addition, the traditional fault detection method mainly relies on simple threshold judgment and lacks intelligent prediction and prevention capabilities, resulting in the camera being unable to switch to the backup power supply in time when a power failure occurs, causing monitoring interruption. In terms of data synchronization and transmission, power fluctuations may cause data loss or transmission delays, affecting the real-time and reliability of the monitoring system.

[0004] In order to address the above problems, the present invention proposes an innovative multi-camera-based power distribution method, which significantly improves the system's power management efficiency, fault detection capability and data transmission stability through intelligent load prediction, autoencoder-based fault detection, efficient data synchronization and transmission monitoring, etc. Summary of the invention

[0005] Based on the above objectives, the present invention provides a power distribution method and system based on multiple cameras.

[0006] The power distribution method based on multiple cameras includes the following steps:

[0007] S1, power distribution system initialization: start the power distribution control module and perform system self-test; initialize the camera connection configuration to ensure that all cameras are normally connected to the power distribution system; conduct a preliminary assessment of the power demand of each camera and generate a preliminary power distribution plan based on the assessment results;

[0008] S2, dynamic power dispatching: real-time monitoring of the working status of each camera, including power consumption and workload; based on the monitoring results and the generated preliminary power distribution plan, dynamically adjust the power distribution of each camera to optimize the power utilization efficiency of the overall power distribution system;

[0009] S3, data synchronization and transmission: establish a data synchronization mechanism between cameras to ensure real-time transmission and processing of image data; the power distribution control module monitors data transmission to prevent data loss or delay caused by power fluctuations;

[0010] S4, fault detection and handling:

[0011] S41, monitor the power supply of each camera in real time to detect potential power failures;

[0012] S42, when a fault is detected, immediately start the backup power module to ensure the continuous operation of the camera;

[0013] S43, record the fault information and remotely notify the administrator to handle it;

[0014] S5, Intelligent Load Forecasting: By analyzing historical power usage data and current working status, the power demand of each camera in the future is predicted; based on the prediction results, the power distribution strategy is adjusted in advance to avoid sudden power shortages or overloads; combined with external data such as weather forecasts, the accuracy of load forecasting is further optimized.

[0015] Furthermore, the S1 specifically includes:

[0016] S11, collecting historical power usage data of each camera, including current, voltage, and power consumption;

[0017] S12, evaluates the working mode and load characteristics of each camera, and analyzes its peak power demand and average power demand;

[0018] S13, determining the priority of power distribution according to the installation location of the camera and the importance of the monitoring area;

[0019] S14, formulating a preliminary power distribution plan based on the total power supply capacity of the power distribution system;

[0020] S15, generating a power distribution plan, including the power distribution ratio of each camera, the power supply time period and the backup power arrangement;

[0021] S16, inputting the power distribution plan into the power distribution control module for real-time management and adjustment.

[0022] Further, the power consumption in S2 includes current, voltage and power consumption data, and the workload includes the speed of processing image data, frame rate, data transmission rate load parameters;

[0023] In S2, the power distribution of each camera is dynamically adjusted according to the monitoring results to optimize the power efficiency of the overall power distribution system, including regularly calculating the deviation between the actual power demand of each camera and the preliminary power distribution plan, which is calculated as follows:

[0024] S21, let: P i,real is the real-time power consumption of the i-th camera, P i,allocis the power distribution of the ith camera in the preliminary power distribution scheme, ΔP i is the power demand deviation of the ith camera, P adj It is an adjustable power resource for the whole system;

[0025] S22, calculate the power demand deviation of each camera: ΔP i =P i,real -P i,alloc ;

[0026] S23, calculate the total power demand deviation of all cameras:

[0027] S24, calculates the dynamic adjustment power of each camera:

[0028] Among them, ΔP total is the sum of the power demand deviations of all cameras, P i,adj Dynamically adjust power for the i-th camera;

[0029] S25, reallocate power according to the dynamic adjustment result: P i,new =P i,alloc +P i,adj , where P i,new The new power allocation after adjustment for the i-th camera.

[0030] Furthermore, the data synchronization mechanism in S3 includes equipping each camera with a time synchronization module, using the network time protocol for time synchronization, and ensuring that the timestamps of all cameras are consistent;

[0031] S31, establish a central data synchronization server to collect, aggregate and distribute the image data of each camera;

[0032] S32, each camera sends the captured image data to the central data synchronization server in chronological order, and the central data synchronization server sorts and synchronizes the data;

[0033] S33 adopts a data cache mechanism. During the data transmission process, a cache queue is used to store the data to be sent and received to prevent data loss.

[0034] Furthermore, the power distribution control module in S3 monitors data transmission specifically including:

[0035] S34, the power distribution control module monitors the power supply status of each camera in real time, especially the power stability during data transmission;

[0036] S35, uses data transmission monitoring algorithms to detect data transmission delay, packet loss rate, and error rate indicators, and evaluates the quality of data transmission in real time;

[0037] S36, automatically switches to backup power when power fluctuations are detected, ensuring continuous operation of the camera and data synchronization server;

[0038] S37, implement a data redundancy mechanism. When a camera's data transmission fails, other cameras temporarily take over part of the task in its monitoring area to ensure the continuity of monitoring data.

[0039] S38, the power distribution control module regularly records and analyzes the data transmission log.

[0040] Furthermore, the fault detection and processing in S4 specifically includes:

[0041] S411, using a current sensor and a voltage sensor to monitor the power supply of each camera in real time, recording current, voltage and power data, and the power distribution control module regularly collects and analyzes the monitoring data;

[0042] S412, introduce a fault prediction model based on an autoencoder to detect potential power distribution faults, use historical monitoring data and real-time data to train the fault prediction model, the autoencoder includes an encoder and a decoder, and identifies anomalies by reconstructing input data. The power supply data of each camera is analyzed in real time and input into the trained autoencoder model for reconstruction. The reconstruction error is calculated. If the error exceeds the preset threshold, it is considered that there is a potential fault. The power distribution control module triggers the fault warning mechanism, records the abnormal information and marks it as a potential fault. When a power distribution fault is detected, the backup power module is immediately started to ensure the continuous operation of the camera.

[0043] Furthermore, the fault prediction model based on the autoencoder is trained as follows:

[0044] The autoencoder consists of an encoder and a decoder. The encoder maps the input data x to a low-dimensional representation z, and the decoder reconstructs z as The encoder function is expressed as: z = f(x), where f is the encoder function, x is the input data, and z is the encoded low-dimensional representation; the decoder function is expressed as: Where g is the decoder function, is the reconstructed data;

[0045] The goal of training an autoencoder is to make the reconstructed data Gradually approach the original input data x, and optimize the parameters of the autoencoder by minimizing the loss function L so that it can reconstruct the input data;

[0046] Real-time fault detection: Transform real-time power supply data intoreal Input the trained fault prediction model to obtain the reconstructed data Calculate the reconstruction error e;

[0047] Fault judgment: Set the reconstruction error threshold T. If e>T, it is considered that there is a potential fault.

[0048] Furthermore, the intelligent load prediction in S5 specifically includes:

[0049] S51, collecting and storing historical power usage data of each camera, including current, voltage, and power consumption, and preprocessing the historical data to remove outliers and noise;

[0050] S52, using improved time series analysis methods to model and predict future electricity demand;

[0051] S53, based on the prediction results, adjust the power distribution strategy in advance, compare the prediction results with the current power supply capacity, determine the time period of power shortage or overload, adjust the power distribution of each camera according to the predicted power demand, give priority to ensuring the power supply of key cameras, if the prediction results show that the power is overloaded in a certain period of time, reduce the power supply of low-priority cameras in advance.

[0052] Furthermore, the step of combining the weather forecast external data in S5 to optimize the accuracy of load prediction specifically includes:

[0053] S54, obtaining weather forecast data, including temperature, humidity, rainfall and wind speed;

[0054] S55, taking weather forecast data as external input, combined with the improved time series model for joint prediction, the joint prediction model is expressed as: containing multiple features W t =[W 1,t ,W 2,t ,W k,t ],W t For weather forecast data:

[0055] in, is the predicted power demand at the future time t+h, W q,t is the weather characteristic data, θ q are model parameters of weather characteristics.

[0057] The multi-camera-based power distribution system is used to implement the multi-camera-based power distribution method, and includes the following modules:

[0058] Power distribution control module: responsible for starting and performing system self-checks to ensure that the system hardware and software functions properly, initializing camera connections, ensuring that all cameras are correctly connected to the power distribution system, and generating a preliminary power distribution plan based on the preliminary power demand assessment results of the cameras;

[0059] Power monitoring and dispatching module: monitors the power consumption and workload of each camera, dynamically adjusts power distribution, optimizes the power efficiency of the overall system, and ensures load balancing;

[0060] Data synchronization and transmission module: establish a data synchronization mechanism to ensure real-time transmission and processing of data between cameras, and monitor the data transmission process;

[0061] Fault detection and emergency handling module: monitors power supply in real time, uses prediction models to predict faults, and activates backup power when power failure is detected;

[0062] Intelligent prediction and optimization module: Analyze historical power usage data and current working status, predict future power demand, adjust the power distribution strategy in advance according to the prediction results, avoid power shortage or overload, and optimize the accuracy of load prediction by combining external factors.

[0063] Beneficial effects of the present invention:

[0064] The present invention uses intelligent load forecasting to analyze historical power usage data and current working status, adopts an improved time series analysis method and autoencoder model, and combines weather forecast data to predict power demand. According to the forecast results, the power distribution strategy is adjusted in advance to prioritize the power supply of key cameras and avoid sudden power shortages or overloads. For example, in the traffic hub monitoring system, it can ensure that the entrance and exit cameras can continue to work stably during peak hours, thereby improving the reliability and operating efficiency of the system. Through this innovation, the system's intelligence in power distribution and its ability to respond to emergencies have been greatly improved.

[0065] The present invention introduces a fault prediction model based on an autoencoder, uses historical and real-time power supply data for training, monitors the power supply status of each camera in real time, and detects potential power distribution faults in a timely manner by calculating the reconstruction error, triggering the fault warning mechanism, and immediately starting the backup power supply module to ensure the continuous operation of the camera. For example, in the monitoring system of a large transportation hub, it can effectively avoid monitoring interruptions caused by power failures and improve the stability and security of the monitoring system. This innovative method greatly improves the accuracy and timeliness of fault detection through advanced machine learning technology.

[0066] The present invention establishes an efficient data synchronization mechanism between the cameras to ensure the real-time transmission and processing of image data. The power distribution control module monitors the data transmission process in real time, and adopts a cache queue and a high-speed data transmission protocol to prevent data loss or delay caused by power fluctuations. When power fluctuations are detected, the system can quickly switch to the backup power supply module to ensure the continuity and stability of data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0068] Figure 1 is a schematic diagram of an embodiment of the present invention;

[0069] Figure 2 The figure is a schematic diagram of the intelligent load prediction process according to an embodiment of the present invention. DETAILED DESCRIPTION

[0070] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0071] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0072] like Figure 1 As shown, the power distribution method based on multiple cameras includes the following steps:

[0073] S1, power distribution system initialization:

[0074] S11, start the power distribution control module and perform system self-check;

[0075] S12, initializing camera connection configuration to ensure that all cameras are properly connected to the power distribution system;

[0076] S13, performing a preliminary evaluation on the power demand of each camera, and generating a preliminary power distribution plan according to the evaluation result;

[0077] S2, dynamic power dispatch:

[0078] S21, real-time monitoring of the working status of each camera, including power consumption and workload;

[0079] S22, dynamically adjusting the power distribution of each camera based on the monitoring results and the generated preliminary power distribution plan to optimize the power utilization efficiency of the overall power distribution system, achieve load balancing, and avoid a camera stopping working due to insufficient power;

[0080] S3, data synchronization and transfer:

[0081] S31, establishing a data synchronization mechanism between cameras to ensure real-time transmission and processing of image data;

[0082] S32, the power distribution control module monitors data transmission to prevent data loss or delay caused by power fluctuations;

[0083] S4, fault detection and handling:

[0084] S41, monitor the power supply of each camera in real time to detect potential power failures;

[0085] S42, when a fault is detected, immediately start the backup power module to ensure the continuous operation of the camera;

[0086] S43, record the fault information and remotely notify the administrator to handle it;

[0087] S5, Intelligent Load Forecasting: By analyzing historical power usage data and current working status, the power demand of each camera in the future is predicted; based on the prediction results, the power distribution strategy is adjusted in advance to avoid sudden power shortages or overloads; combined with external data such as weather forecasts, the accuracy of load forecasting is further optimized.

[0088] S11 is as follows:

[0089] a. Start the power distribution control module and activate its internal processing unit and control logic;

[0090] b. System self-check includes the following:

[0091] Check whether the output voltage and current of the power module are within the normal range;

[0092] Check whether the communication module can communicate normally with all cameras;

[0093] Check whether the hardware and software functions of the power distribution control module are operating normally;

[0094] Check whether the backup power module is available;

[0095] c. Generate a status report based on the self-test results to confirm whether the power distribution system is ready.

[0096] S12 is as follows:

[0097] a. Check the connection status of all cameras to ensure that the physical connection is correct;

[0098] b. Confirm that each camera’s unique identifier (ID) matches the configuration file in the power distribution system;

[0099] c. Configure the communication parameters of each camera, including communication protocol and transmission rate;

[0100] d. Verify whether each camera can receive and respond to the instructions of the power distribution control module;

[0101] e. After the camera is successfully connected to the power distribution system, record the camera's connection status and parameter information.

[0102] S13 specifically includes:

[0103] S131, collecting historical power usage data of each camera, including current, voltage and power consumption;

[0104] S132, evaluating the working mode and load characteristics of each camera, and analyzing its peak power demand and average power demand;

[0105] S133, determining the priority of power distribution according to the installation location of the camera and the importance of the monitoring area;

[0106] S134, formulate a preliminary power distribution plan based on the total power supply capacity of the power distribution system;

[0107] S135, generating a power distribution plan, including the power distribution ratio of each camera, the power supply time period and the backup power arrangement;

[0108] S136, input the power distribution plan into the power distribution control module for real-time management and adjustment.

[0109] The calculation formula of the power distribution scheme is as follows, assuming:

[0110] P i is the peak power demand of the i-th camera;

[0111] Ai is the importance weight of the working mode of the i-th camera (dimensionless);

[0112] T i is the working time of the i-th camera;

[0113] E is the total power supply of the system;

[0114] n is the total number of cameras;

[0115] Calculation steps:

[0116] 1. Calculate the weighted power requirement of each camera: W i =P i ×A i ;

[0117] 2. Calculate the sum of weighted power requirements of all cameras:

[0118] 3. Calculate the power allocation ratio for each camera:

[0119] 4. Calculate the power distribution of each camera based on the total power supply and working time:

[0120] E i =R i ×E;

[0121] Among them, W i is the weighted power requirement of the ith camera, W total is the sum of the weighted power requirements of all cameras, R i is the power allocation ratio of the i-th camera, E i is the actual power allocation of the i-th camera. Through the above calculation, the power demand of each camera can be reasonably allocated to ensure that the cameras in key areas are given priority in power supply, while ensuring the power efficiency of the overall system.

[0122] The power consumption in S21 includes current, voltage and power consumption data, and the workload includes the speed of processing image data, frame rate, and data transmission rate load parameters;

[0123] In S22, the power distribution of each camera is dynamically adjusted according to the monitoring results to optimize the power efficiency of the overall power distribution system, including periodically (such as every minute) calculating the deviation between the actual power demand of each camera and the preliminary power distribution plan, as follows:

[0124] S221, let: P i,real is the real-time power consumption of the i-th camera, P i,alloc is the power distribution of the ith camera in the preliminary power distribution scheme, ΔPi is the power demand deviation of the ith camera, P adj It is an adjustable power resource for the whole system;

[0125] S222, calculate the power demand deviation of each camera: ΔP i =P i,real -P i,alloc ;

[0126] S223, calculating the total power demand deviation of all cameras:

[0127] S224, calculating the dynamic adjustment power of each camera:

[0128] Among them, ΔP total is the sum of the power demand deviations of all cameras, P i,adj Dynamically adjust power for the i-th camera;

[0129] S225, reallocate power according to the dynamic adjustment result: P i,new =P i,alloc +P i,adj , where P i,new The new power allocation after adjustment for the i-th camera.

[0130] Ensure load balancing of the overall system:

[0131] When adjusting power distribution, give priority to ensuring the power needs of key cameras;

[0132] For cameras with excess power, their power allocation will be appropriately reduced, and the excess power will be allocated to cameras with higher demand;

[0133] Update the power distribution table in real time to ensure the working stability and continuity of all cameras.

[0134] The data synchronization mechanism in S31 includes equipping each camera with a time synchronization module, using the network time protocol for time synchronization to ensure that the timestamps of all cameras are consistent;

[0135] S311, establish a central data synchronization server to collect, aggregate and distribute image data from each camera;

[0136] S312, each camera sends the captured image data to the central data synchronization server in chronological order, and the central data synchronization server sorts and synchronizes the data;

[0137] S313, adopting a data cache mechanism, during the data transmission process, using a cache queue to store the data to be sent and received to prevent data loss.

[0138] 1. Time synchronization calculation:

[0139] Use Network Time Protocol (NTP) for time synchronization: Set T i Indicates the local time of camera i, T s Indicates the time of the standard time server;

[0140] Time synchronization deviation calculation: ΔT i =T i -T s , where ΔT i is the time deviation of camera i;

[0141] Adjust local time: T i ′=T i -ΔT i , where T i ' is the adjusted local time;

[0142] 2. Data caching mechanism:

[0143] Data cache queue: Assume the length of the data cache queue is L and the transmission rate is R;

[0144] Data cache queue length calculation: L = R × T buffer , where T buffer The cache time.

[0145] The power distribution control module in S32 monitors data transmission specifically including:

[0146] S321, the power distribution control module monitors the power supply status of each camera in real time, especially the power stability during data transmission;

[0147] S322, using a data transmission monitoring algorithm to detect data transmission delay, packet loss rate, and error rate indicators, and evaluate the quality of data transmission in real time;

[0148] S323, automatically switches to backup power when power fluctuations are detected to ensure continuous operation of the camera and data synchronization server;

[0149] S323, implement a data redundancy mechanism. When a camera's data transmission fails, other cameras temporarily take over part of the task of monitoring the area to ensure the continuity of monitoring data.

[0150] S324, the power distribution control module regularly records and analyzes the data transmission log to discover and solve potential problems and improve the reliability of the system.

[0151] The delay is calculated as follows:

[0152] Packet transmission delay: Let Tsend is the packet sending time, T receive Data packet receiving time: D = T receive -T send , where D is the transmission delay of the data packet.

[0153] Packet loss rate calculation: Assume that a total of N packets are sent and M packets are received:

[0154] Among them, P loss is the packet loss rate.

[0155] Packet error rate: Assume there are E erroneous packets among the sent packets:

[0156] Among them, P error is the error rate.

[0157] Power fluctuation detection: Assume that the real-time power supply is P current , the power fluctuation threshold is set to P threshold :ΔP=|P current -P expected |, where P expected for the desired power supply.

[0158] Backup power switch: When power fluctuation is detected, the backup power module is activated:

[0159] P backup =P current .

[0160] Fault detection and processing in S4 specifically include:

[0161] S411, using a current sensor and a voltage sensor to monitor the power supply of each camera in real time, recording current, voltage and power data, and the power distribution control module regularly (every second) collects and analyzes the monitoring data;

[0162] S412, introduce a fault prediction model based on an autoencoder to detect potential power distribution faults, use historical monitoring data and real-time data to train the fault prediction model, the autoencoder includes an encoder and a decoder, and identifies anomalies by reconstructing input data. The power supply data of each camera is analyzed in real time and input into the trained autoencoder model for reconstruction. The reconstruction error is calculated. If the error exceeds a preset threshold, it is considered that there is a potential fault. The power distribution control module triggers a fault warning mechanism, records the abnormal information and marks it as a potential fault. When a power distribution fault is detected, the backup power module is immediately started to ensure the continuous operation of the camera:

[0163] The power distribution control module records the time, cause and treatment measures of the fault, and remotely notifies the system administrator for further inspection and maintenance. After the fault is eliminated, the power distribution control module switches the system power supply back to the main power supply, and charges and maintains the backup power supply module for next use.

[0164] The fault prediction model based on the autoencoder is trained as follows:

[0165] The autoencoder consists of an encoder and a decoder. The encoder maps the input data x to a low-dimensional representation z, and the decoder reconstructs z as The encoder function is expressed as: z = f(x), where f is the encoder function, x is the input data, and z is the encoded low-dimensional representation; the decoder function is expressed as: Where g is the decoder function, is the reconstructed data;

[0166] The goal of training an autoencoder is to make the reconstructed data Gradually approaching the original input data x, the loss function is: in, is the sum of squares of the reconstruction error. By minimizing the loss function L, the parameters of the autoencoder are optimized so that it can reconstruct the input data.

[0167] Real-time fault detection: Transform real-time power supply data into real Input the trained fault prediction model to obtain the reconstructed data Reconstruction error calculation: Where, e is the reconstruction error;

[0168] Fault judgment: Set the reconstruction error threshold T. If e>T, it is considered that there is a potential fault.

[0169] In the multi-camera-based power distribution method of the present invention, the specific steps of using the autoencoder to detect power distribution faults are as follows:

[0170] 1. Data collection and preprocessing:

[0171] a. Collect power supply data x in real time from the current sensor and voltage sensor of each camera, including current I, voltage V and power P.

[0172] b. Standardize the collected data to meet the input requirements of the autoencoder.

[0173] 2. Model training:

[0174] a. Use historical data to train the autoencoder model, including data samples of normal and abnormal states.

[0175] b. Encode and decode the training data to optimize the model parameters by minimizing the reconstruction error.

[0176] c. Verify the reconstruction capability of the model to ensure that the model can accurately reconstruct the data in the normal state.

[0177] 3. Real-time fault detection:

[0178] a. Collect the real-time power supply data x real Input the autoencoder model for reconstruction to obtain reconstructed data

[0179] b. Calculate the reconstruction error e and compare it with the preset threshold T: If e>T, a fault warning is triggered.

[0180] Smart load forecasting in S5 specifically includes:

[0181] S51, collecting and storing historical power usage data of each camera, including current, voltage, and power consumption, and preprocessing the historical data to remove outliers and noise;

[0182] S52, using the improved time series analysis method to build a model and predict future power demand, assuming that the historical power data of camera i is P i,t :

[0183] in, is the predicted power demand at the future time t+h, P i,t-j+k is the actual power demand at the past time, t-j+k, F i,t-l is the current working status characteristics (such as the speed of processing image data, frame rate, etc.), γ k and δ l is the model parameter, ∈ t is the prediction error;

[0184] S53, based on the prediction results, the power distribution strategy is adjusted in advance, the prediction results are compared with the current power supply capacity, the power shortage or overload time period is determined, and the power distribution of each camera is adjusted according to the predicted power demand, and the power supply of key cameras is prioritized. If the prediction results show that the power is overloaded during a certain period, the power supply of low-priority cameras is reduced in advance to ensure the stability of the overall system. Key cameras include entrance and exit surveillance cameras, bank or vault surveillance cameras, transportation hub surveillance cameras, and important equipment surveillance cameras.

[0185] S5 combines weather forecast external data to optimize the accuracy of load forecasting, including:

[0186] S54, obtaining weather forecast data, including temperature, humidity, rainfall and wind speed;

[0187] S55, taking weather forecast data as external input, combined with the improved time series model for joint prediction, the joint prediction model is expressed as: containing multiple features W t =[W 1,t ,W 2,t ,W k,t ],W t For weather forecast data:

[0188] in, is the predicted power demand at the future time t+h, W q,t is the weather characteristic data, θ q are model parameters of weather characteristics.

[0190] like Figure 2 As shown, the multi-camera-based power distribution system is used to implement the multi-camera-based power distribution method, and includes the following modules:

[0191] Power distribution control module: responsible for starting and performing system self-checks to ensure that the system hardware and software functions properly, initializing camera connections, ensuring that all cameras are correctly connected to the power distribution system, and generating a preliminary power distribution plan based on the preliminary power demand assessment results of the cameras;

[0192] Power monitoring and dispatching module: monitors the power consumption and workload of each camera, dynamically adjusts power distribution, optimizes the power efficiency of the overall system, and ensures load balancing;

[0193] Data synchronization and transmission module: establish a data synchronization mechanism to ensure real-time transmission and processing of data between cameras, and monitor the data transmission process;

[0194] Fault detection and emergency handling module: monitors power supply in real time, uses prediction models to predict faults, and activates backup power when power failure is detected;

[0195] Intelligent prediction and optimization module: Analyze historical power usage data and current working status, predict future power demand, adjust the power distribution strategy in advance according to the prediction results, avoid power shortage or overload, and optimize the accuracy of load prediction by combining external factors.

[0196] Those skilled in the art should understand that the discussion of any of the above embodiments is merely illustrative and is not intended to imply that the scope of the present invention is limited to these examples. Under the concept of the present invention, the technical features in the above embodiments or different embodiments may be combined, the steps may be implemented in any order, and there are many other variations of the different aspects of the present invention as described above, which are not provided in detail for the sake of simplicity.

Claims

1. A power distribution method based on multiple cameras, characterized in that: The following steps are involved: S1, power distribution system initialization: start the power distribution control module and perform system self-test; initialize the camera connection configuration to ensure that all cameras are normally connected to the power distribution system; conduct a preliminary assessment of the power demand of each camera and generate a preliminary power distribution plan based on the assessment results; S2, dynamic power dispatching: real-time monitoring of the working status of each camera, including power consumption and workload; based on the monitoring results and the generated preliminary power distribution plan, dynamically adjust the power distribution of each camera to optimize the power utilization efficiency of the overall power distribution system; S3, data synchronization and transmission: establish a data synchronization mechanism between cameras to ensure real-time transmission and processing of image data; the power distribution control module monitors data transmission to prevent data loss or delay caused by power fluctuations; S4, fault detection and handling: S41, monitor the power supply of each camera in real time to detect potential power failures; S42, when a fault is detected, immediately start the backup power module to ensure the continuous operation of the camera; S43, record the fault information and remotely notify the administrator to handle it; S5, Intelligent Load Forecasting: By analyzing historical power usage data and current working status, the power demand of each camera in the future is predicted; based on the prediction results, the power distribution strategy is adjusted in advance to avoid sudden power shortages or overloads; combined with external data such as weather forecasts, the accuracy of load forecasting is further optimized.

2. The multi-camera-based power distribution method according to claim 1, characterized in that: The S1 specifically includes: S11, collecting historical power usage data of each camera, including current, voltage, and power consumption; S12, evaluates the working mode and load characteristics of each camera, and analyzes its peak power demand and average power demand; S13, determining the priority of power distribution according to the installation location of the camera and the importance of the monitoring area; S14, formulating a preliminary power distribution plan based on the total power supply capacity of the power distribution system; S15, generating a power distribution plan, including the power distribution ratio of each camera, the power supply time period and the backup power arrangement; S16, inputting the power distribution plan into the power distribution control module for real-time management and adjustment.

3. The multi-camera-based power distribution method according to claim 1, characterized in that: The power consumption in S2 includes current, voltage and power consumption data, and the workload includes the speed of processing image data, frame rate, and data transmission rate load parameters; In S2, the power distribution of each camera is dynamically adjusted according to the monitoring results to optimize the power efficiency of the overall power distribution system, including regularly calculating the deviation between the actual power demand of each camera and the preliminary power distribution plan, which is calculated as follows: S21, let: P i,real is the real-time power consumption of the i-th camera, P i,alloc is the power distribution of the ith camera in the preliminary power distribution scheme, ΔP i is the power demand deviation of the ith camera, P adj It is an adjustable power resource for the whole system; S22, calculate the power demand deviation of each camera: ΔP i =P i,real -P i,alloc ; S23, calculate the total power demand deviation of all cameras: S24, calculates the dynamic adjustment power of each camera: Among them, ΔP total is the sum of the power demand deviations of all cameras, P i,adj Dynamically adjust power for the i-th camera; S25, reallocate power according to the dynamic adjustment result: P i,new =P i,alloc +P i,adj , where P i,new The new power allocation after adjustment for the i-th camera.

4. The multi-camera-based power distribution method according to claim 1, characterized in that: The data synchronization mechanism in S3 includes equipping each camera with a time synchronization module, using the network time protocol for time synchronization to ensure that the timestamps of all cameras are consistent; S31, establish a central data synchronization server to collect, aggregate and distribute the image data of each camera; S32, each camera sends the captured image data to the central data synchronization server in chronological order, and the central data synchronization server sorts and synchronizes the data; S33 adopts a data cache mechanism. During the data transmission process, a cache queue is used to store the data to be sent and received to prevent data loss.

5. The multi-camera-based power distribution method according to claim 4, characterized in that: The power distribution control module in S3 monitors data transmission specifically including: S34, the power distribution control module monitors the power supply status of each camera in real time, especially the power stability during data transmission; S35, uses a data transmission monitoring algorithm to detect data transmission delay, packet loss rate, and error rate indicators, and evaluates the quality of data transmission in real time; S36, automatically switches to backup power when power fluctuations are detected, ensuring continuous operation of the camera and data synchronization server; S37, implement data redundancy mechanism. When a camera's data transmission fails, other cameras temporarily take over part of the task of monitoring the area to ensure the continuity of monitoring data. S38, the power distribution control module regularly records and analyzes the data transmission log.

6. The multi-camera-based power distribution method according to claim 1, characterized in that: The fault detection and processing in S4 specifically includes: S411, using a current sensor and a voltage sensor to monitor the power supply of each camera in real time, recording current, voltage and power data, and the power distribution control module regularly collects and analyzes the monitoring data; S412, introduce a fault prediction model based on an autoencoder to detect potential power distribution faults, use historical monitoring data and real-time data to train the fault prediction model, the autoencoder includes an encoder and a decoder, and identifies anomalies by reconstructing input data. The power supply data of each camera is analyzed in real time and input into the trained autoencoder model for reconstruction. The reconstruction error is calculated. If the error exceeds the preset threshold, it is considered that there is a potential fault. The power distribution control module triggers the fault warning mechanism, records the abnormal information and marks it as a potential fault. When a power distribution fault is detected, the backup power module is immediately started to ensure the continuous operation of the camera.

7. The multi-camera-based power distribution method according to claim 6, characterized in that: The fault prediction model based on the autoencoder is trained as follows: The autoencoder consists of an encoder and a decoder. The encoder maps the input data x to a low-dimensional representation z, and the decoder reconstructs z as The encoder function is expressed as: z = f(x), where f is the encoder function, x is the input data, and z is the encoded low-dimensional representation; the decoder function is expressed as: where g is the decoder function, is the reconstructed data; The goal of training an autoencoder is to make the reconstructed data Gradually approach the original input data x, and optimize the parameters of the autoencoder by minimizing the loss function L so that it can reconstruct the input data; Real-time fault detection: Transform real-time power supply data into real Input the trained fault prediction model to obtain the reconstructed data Calculate the reconstruction error e; Fault judgment: Set the reconstruction error threshold T. If e>T, it is considered that there is a potential fault.

8. The multi-camera-based power distribution method according to claim 1, characterized in that: The intelligent load prediction in S5 specifically includes: S51, collecting and storing historical power usage data of each camera, including current, voltage, and power consumption, and preprocessing the historical data to remove outliers and noise; S52, using improved time series analysis methods to model and predict future electricity demand; S53, based on the prediction results, adjust the power distribution strategy in advance, compare the prediction results with the current power supply capacity, determine the time period of power shortage or overload, adjust the power distribution of each camera according to the predicted power demand, give priority to ensuring the power supply of key cameras, if the prediction results show that the power is overloaded in a certain period of time, reduce the power supply of low-priority cameras in advance.

9. The multi-camera-based power distribution method according to claim 8, characterized in that: The step of combining weather forecast external data in S5 to optimize the accuracy of load prediction specifically includes: S54, obtaining weather forecast data, including temperature, humidity, rainfall and wind speed; S55, takes weather forecast data as external input and combines it with the improved time series model for joint prediction.

10. A multi-camera based power distribution system, used to implement the multi-camera based power distribution method according to any one of claims 1 to 9, characterized in that: Includes the following modules: Power distribution control module: responsible for starting and performing system self-checks to ensure that the system hardware and software functions properly, initializing camera connections, ensuring that all cameras are correctly connected to the power distribution system, and generating a preliminary power distribution plan based on the preliminary power demand assessment results of the cameras; Power monitoring and dispatching module: monitors the power consumption and workload of each camera, dynamically adjusts power distribution, optimizes the power efficiency of the overall system, and ensures load balancing; Data synchronization and transmission module: establish a data synchronization mechanism to ensure real-time transmission and processing of data between cameras, and monitor the data transmission process; Fault detection and emergency handling module: monitors power supply in real time, uses prediction models to predict faults, and activates backup power when power failure is detected; Intelligent prediction and optimization module: Analyze historical power usage data and current working status, predict future power demand, adjust the power distribution strategy in advance according to the prediction results, avoid power shortage or overload, and optimize the accuracy of load prediction by combining external factors.

Images