Intelligent security risk assessment method and system based on multi-source data fusion

Abstract

The invention discloses an intelligent security risk assessment method and system based on multi-source data fusion, and relates to the technical field of security, and the method comprises the steps: collecting environment data in real time through deploying a sensor network, carrying out the preprocessing, and inputting the environment data into a risk assessment model; multi-source data features are extracted through a machine learning algorithm to divide risk levels, and a risk assessment result is updated in real time; and dynamically adjusting the operation state of the sensor according to the risk level, and compensating the monitoring blind area through a data fusion algorithm. According to the method, through multi-source data fusion, the monitoring blind area of a single sensor is made up, and the comprehensiveness and accuracy of risk detection are improved; the operation state of the sensor is dynamically adjusted according to the risk level, unnecessary sensors are closed when the risk is low, energy consumption is saved, and the service life of equipment is prolonged; by triggering an alarm upgrading mechanism, the risk level is improved according to historical records and continuous evaluation results, it is ensured that stricter security measures can be taken in time, and accidents are avoided.

Description

Technical Field

[0001] The present invention relates to the field of security technology, and in particular to an intelligent security risk assessment method and system based on multi-source data fusion. Background Art

[0002] In recent years, with the rapid development of information technology and intelligent equipment, intelligent security systems have gradually become an important means of ensuring public safety. Traditional security systems mostly rely on single sensors or simple data processing. With the increase in security needs, multi-source data fusion technology has emerged. Multi-source data fusion can comprehensively utilize data from multiple devices such as cameras, radars, infrared sensors, etc. to improve the accuracy and reliability of risk detection. However, existing technologies still have shortcomings in data processing efficiency, real-time performance and energy consumption management.

[0003] The existing intelligent security risk assessment methods have limited real-time processing capabilities for multi-source data, making it difficult to meet the rapid response requirements in high-risk scenarios; the data fusion algorithm is relatively simple and lacks effective compensation for monitoring blind spots, resulting in poor monitoring effects; the sensor's energy consumption management is not intelligent enough, resulting in a waste of resources, and in high-risk situations, it may cause energy consumption to exceed the safety threshold, affecting system stability. Based on the above problems, the present invention effectively solves these problems by introducing an efficient data preprocessing method, an intelligent data fusion algorithm, and a dynamically adjusted sensor management strategy, thereby improving the real-time performance and reliability of the system. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problems solved by the present invention are: the existing technology has low data processing efficiency, insufficient compensation for monitoring blind spots, and insufficient intelligent energy consumption management, as well as how to achieve efficient real-time processing and intelligent fusion of multi-source data.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: an intelligent security risk assessment method based on multi-source data fusion, comprising: deploying a sensor network to collect environmental data in real time and pre-processing it, and inputting the environmental data into a risk assessment model; extracting multi-source data features through a machine learning algorithm to divide the risk level, and updating the risk assessment results in real time; dynamically adjusting the sensor operating status according to the risk level, and compensating for monitoring blind spots through a data fusion algorithm.

[0007] As a preferred solution of the intelligent security risk assessment method based on multi-source data fusion described in the present invention, the sensor network includes cameras, radars, infrared sensors, temperature and humidity sensors, and sound sensors.

[0008] As a preferred solution of the intelligent security risk assessment method based on multi-source data fusion described in the present invention, the environmental data collection and pre-processing includes equipping each sensor with an independent data acquisition module, and transmitting the collected data to the central processing unit or edge computing node in real time through the 5G network; ensuring the security of data transmission through an encryption protocol, and compressing the data through the LZ77 algorithm; denoising the original data, unifying the format, and synchronously sampling the timestamp with the NTP protocol to generate a multi-source data set with a unified time axis for input into the risk assessment model.

[0009] As a preferred solution of the intelligent security risk assessment method based on multi-source data fusion described in the present invention, the risk assessment model includes building a risk assessment model based on a deep neural network; the input layer receives multi-source data, the hidden layer includes 3 fully connected layers, and the activation function is ReLU; features are extracted from the multi-source data, and the output layer divides the risk assessment level by the Softmax function; the risk assessment level includes low risk c1, medium risk c2 and high risk c3; the risk level probability distribution is generated by the Softmax function, and the risk level probability is expressed as:

[0010]

[0011] Among them, P(c) represents the probability of the risk level of category c, the output range is [0,1], the sum of the probabilities is 1, z c Represents the unnormalized model prediction value of risk level c; if P(c)>0.8, it is judged as low risk; if 0.5<P(c)≤0.8, it is judged as medium risk; if P(c)≤0.5, it is judged as high risk; the assessment result is updated every 200ms.

[0012] As a preferred solution of the intelligent security risk assessment method based on multi-source data fusion described in the present invention, the real-time update of risk assessment results includes deploying a risk assessment model on the edge computing node. When the sensor network data is input, the real-time computing thread is triggered, and the computing thread priority is set to the highest. If the computing resources of the edge node are insufficient, the data is forwarded to the cloud server through the message queue, and the cloud server uses a distributed computing framework to process multi-node data in parallel. The assessment results are pushed to the central processing unit through the MQTT protocol and compared with the event records in the historical database. If the assessment is high risk for three consecutive times and there are similar intrusion events in the historical records, the alarm upgrade mechanism is triggered and the risk level is increased by one level.

[0013] As a preferred solution of the intelligent security risk assessment method based on multi-source data fusion described in the present invention, the dynamic adjustment of the sensor operating status according to the risk level includes: if the risk level is assessed as low risk, turning off the camera and sound sensor, retaining the radar, infrared sensor and temperature and humidity sensor; if the risk level is assessed as medium risk, turning on 50% of the cameras according to the alternating polling strategy, and starting the sound sensor according to the threshold; if the risk level is assessed as high risk, turning on all sensors, switching the camera to high frame rate mode, increasing the radar scanning frequency to 10Hz, and enabling the directional adaptation function of the sound sensor; when the sensor state is switched, the controller verifies the current energy consumption. If the total power consumption exceeds the safety threshold, the non-critical sensors are forcibly turned off according to the priority; the sensor shutdown and startup instructions are sent to the sensor controller via the Modbus protocol.

[0014] As a preferred solution of the intelligent security risk assessment method based on multi-source data fusion described in the present invention, wherein: the compensation of the monitoring blind spot by the data fusion algorithm includes starting the data fusion algorithm when the camera is turned off; the data fusion algorithm includes a radar and infrared complementary strategy and a redundancy elimination strategy; the radar and infrared complementary strategy includes the coordinates of the moving object detected by the radar (x r ,y r ) and infrared heat source coordinates (x i ,y i ) for spatial matching, if |x r -x i |≤0.5m and |y r -y i |≤0.5m, the moving objects are determined to be the same target, and the target threat index T is calculated, which is expressed as:

[0015] T=αv+βΔT

[0016] Where v represents the moving speed of the moving object, ΔT represents the temperature difference of the heat source, α and β represent the weight coefficients, α = 0.6 and β = 0.4 respectively. The redundancy elimination strategy includes using the Hungarian algorithm to perform optimal matching on targets detected repeatedly by multiple sensors and retaining the data with the highest confidence. The confidence level C is expressed as:

[0017]

[0018] Among them, d k represents the kth sensor measurement value, μ represents the mean, σ represents the standard deviation, and n represents the total number of sensors.

[0019] Another object of the present invention is to provide an intelligent security risk assessment system based on multi-source data fusion, which can dynamically adjust the operating status of the sensor network through a deep neural network risk assessment model, solving the problem that the current technology is not intelligent enough in sensor energy consumption management and has high energy consumption.

[0020] As a preferred solution of the intelligent security risk assessment system based on multi-source data fusion described in the present invention, it includes: a data processing module, a risk assessment module, and a dynamic adjustment module; the data processing module is used to collect environmental data in real time and pre-process it by deploying a sensor network, and input the environmental data into a risk assessment model; the risk assessment module is used to extract multi-source data features through a machine learning algorithm to divide the risk level and update the risk assessment results in real time; the dynamic adjustment module is used to dynamically adjust the sensor operating status according to the risk level and compensate for the monitoring blind spot through the data fusion algorithm.

[0021] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a step of an intelligent security risk assessment method based on multi-source data fusion.

[0022] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of an intelligent security risk assessment method based on multi-source data fusion.

[0023] Beneficial effects of the present invention: The intelligent security risk assessment method based on multi-source data fusion provided by the present invention compensates for the monitoring blind spots of a single sensor through multi-source data fusion, and improves the comprehensiveness and accuracy of risk detection; by dynamically adjusting the sensor operating status according to the risk level, unnecessary sensors are turned off when the risk is low, saving energy consumption and extending the life of the equipment; by adopting edge computing and cloud distributed processing, it ensures that data can still be processed quickly when computing resources are insufficient, meeting high real-time requirements; by using encryption protocols and compression algorithms, the security and integrity of data during transmission are ensured; the risk assessment model based on deep neural networks updates the assessment results every 200ms, provides timely risk feedback, improves system reliability, and is applicable to real-time monitoring scenarios; through redundancy elimination strategies and complementary strategies, the data processing process is optimized, repeated detection is reduced, and the system operation efficiency is improved while saving energy consumption; by triggering the alarm upgrade mechanism, the risk level is increased according to historical records and continuous assessment results, ensuring that more stringent security measures can be taken in time to avoid accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 This is an overall flow chart of an intelligent security risk assessment method based on multi-source data fusion provided by the first embodiment of the present invention.

[0026] Figure 2 This is an overall flow chart of an intelligent security risk assessment system based on multi-source data fusion provided in the third embodiment of the present invention. DETAILED DESCRIPTION

[0027] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0028] Example 1, with reference to Figure 1 , as one embodiment of the present invention, provides an intelligent security risk assessment method based on multi-source data fusion, comprising:

[0029] S1: Deploy a sensor network to collect environmental data in real time and pre-process it, and then input the environmental data into the risk assessment model.

[0030] Furthermore, the sensor network includes cameras, radars, infrared sensors, temperature and humidity sensors, and sound sensors.

[0031] It should be noted that collecting environmental data and preprocessing it include equipping each sensor with an independent data acquisition module, transmitting the collected data in real time to the central processing unit or edge computing node through the 5G network; ensuring data transmission security through encryption protocols, and compressing the data through the LZ77 algorithm; denoising the original data, unifying the format, and timestamp synchronization sampling NTP protocol to ensure that the time error of each sensor data is less than 10ms, and generating a multi-source data set with a unified time axis for input into the risk assessment model.

[0032] It should also be noted that denoising the raw data includes but is not limited to using Gaussian filtering on the camera video stream to eliminate ambient light interference. The formula is:

[0033]

[0034] Among them, I(x I ,y I ) represents the original pixel value, and σ represents the standard deviation of the Gaussian kernel. The radar data is subjected to Kalman filtering to eliminate motion noise, and the sound sensor data is subjected to wavelet transform to remove low-frequency noise.

[0035] It should also be noted that the unified format includes but is not limited to converting video streams into H.264 encoding and radar data into JSON format.

[0036] It should also be noted that each sensor is equipped with an independent data acquisition module to ensure the independence and reliability of data acquisition; the collected data is transmitted to the central processing unit or edge computing node in real time through the 5G network to ensure the real-time and efficiency of data transmission; the encryption protocol is used to ensure data transmission security to prevent data leakage or tampering; the data is compressed using the LZ777 algorithm to reduce the amount of data transmitted and the storage space occupied; the raw data is denoised to remove noise and interference in the data and improve data quality; the data format collected by different sensors is unified to facilitate subsequent data processing and analysis; the sampling timestamps are synchronized through the NTP protocol to ensure that the data collected by different sensors are aligned in time, which facilitates subsequent data fusion and analysis.

[0037] It should also be noted that deploying sensor networks to collect environmental data in real time and preprocessing it, and inputting environmental data into the risk assessment model is the basis of the entire intelligent security assessment method. By collecting and preprocessing multi-source data in real time, a reliable data foundation is provided for subsequent risk assessment, ensuring the accuracy and real-time nature of risk assessment.

[0038] S2: Extract multi-source data features through machine learning algorithms to divide risk levels and update risk assessment results in real time.

[0039] Furthermore, the risk assessment model includes building a risk assessment model based on a deep neural network; the input layer receives multi-source data, the hidden layer contains three fully connected layers, and the activation function is ReLU; features are extracted from the multi-source data, and the output layer divides the risk assessment level using the Softmax function; the risk assessment level includes low risk c1, medium risk c2, and high risk c3; the Softmax function is used to generate the risk level probability distribution, and the risk level probability is expressed as:

[0040]

[0041] Among them, P(c) represents the probability of the risk level of category c, the output range is [0,1], the sum of the probabilities is 1, z cRepresents the unnormalized model prediction value of risk level c; if P(c)>0.8, it is judged as low risk; if 0.5<P(c)≤0.8, it is judged as medium risk; if P(c)≤0.5, it is judged as high risk; the assessment result is updated every 200ms.

[0042] It should be noted that real-time updating of risk assessment results includes deploying risk assessment models on edge computing nodes. When sensor network data is input, the real-time computing thread is triggered, and the computing thread priority is set to the highest. If the computing resources of the edge node are insufficient, the data is forwarded to the cloud server through the message queue. The cloud server uses a distributed computing framework to process multi-node data in parallel. The assessment results are pushed to the central processing unit through the MQTT protocol and compared with the event records in the historical database. If the assessment is high risk for three consecutive times and there are similar intrusion events in the historical records, the alarm upgrade mechanism is triggered and the risk level is increased by one level.

[0043] It should also be noted that multi-source data includes but is not limited to sensor data, historical security record data, time information data, and weather data.

[0044] It should also be noted that extracting features from multi-source data includes but is not limited to extracting the Euclidean distance change rate of the moving object trajectory from camera data, which is expressed as:

[0045]

[0046] Among them, (x t ,y t ) represents the trajectory coordinates of the moving object in the camera; the speed and acceleration of the moving object are extracted from the radar data; and the Mel-frequency cepstral coefficients MFCC are extracted from the sound sensor as abnormal features.

[0047] It should also be noted that features are extracted from multi-source data through machine learning algorithms to effectively reflect potential risks in the environment; the extracted features are input into the risk assessment model, and the model is used for calculation and analysis to divide the risks into different levels, which refines the response strategies of intelligent security for dealing with different risks, improves the adaptability of security and saves energy; the risk assessment model built based on deep neural networks can effectively identify and assess potential risks in the environment; the risk assessment model updates the assessment results every 200ms to ensure the real-time nature of risk assessment, promptly reflect changes in the environment, and take corresponding security measures based on the latest assessment results; deploying the risk assessment model through edge computing nodes can effectively improve data processing efficiency and meet high real-time requirements; the assessment results are pushed to the central processing unit through the MQTT protocol, and compared with the event records in the historical database to effectively avoid false alarms and omissions, and improve the reliability of the system.

[0048] It should also be noted that extracting multi-source data features through machine learning algorithms and dividing risk levels, and updating risk assessment results in real time are the core of the entire intelligent security risk assessment method. Extracting multi-source data features through machine learning algorithms and dividing risk levels provide a basis for subsequent dynamic adjustments and security measures, ensuring the intelligence and security of the system.

[0049] S3: Dynamically adjust the sensor operating status according to the risk level and compensate for monitoring blind spots through data fusion algorithms.

[0050] Furthermore, the sensor operating status is dynamically adjusted according to the risk level. If the risk level is assessed as low risk, the camera and sound sensor are turned off, and the radar, infrared sensor, and temperature and humidity sensor are retained; if the risk level is assessed as medium risk, 50% of the cameras are turned on according to the alternating polling strategy, and the sound sensor is started according to the threshold; if the risk level is assessed as high risk, all sensors are turned on, the camera is switched to high frame rate mode, the radar scanning frequency is increased to 10Hz, and the sound sensor enables the directional adaptation function; when the sensor state is switched, the controller verifies the current energy consumption. If the total power consumption exceeds the safety threshold, non-critical sensors are forcibly turned off according to the priority; the sensor off and on instructions are sent to the sensor controller via the Modbus protocol.

[0051] It should be noted that the compensation of monitoring blind spots by the data fusion algorithm includes starting the data fusion algorithm when the camera is turned off; the data fusion algorithm includes radar and infrared complementary strategy and redundancy elimination strategy; the radar and infrared complementary strategy includes the coordinates of the moving object detected by the radar (x r ,y r ) and infrared heat source coordinates (x i ,y i ) for spatial matching, if |x r -x i |≤0.5m and |y r -y i |≤0.5m, the moving objects are determined to be the same target, and the target threat index T is calculated, which is expressed as:

[0052] T=αv+βΔT

[0053] Where v represents the moving speed of the moving object, ΔT represents the temperature difference of the heat source, α and β represent the weight coefficients, α = 0.6 and β = 0.4 respectively. The redundancy elimination strategy includes using the Hungarian algorithm to perform optimal matching on targets detected repeatedly by multiple sensors and retaining the data with the highest confidence. The confidence level C is expressed as:

[0054]

[0055] Among them, d krepresents the kth sensor measurement value, μ represents the mean, σ represents the standard deviation, and n represents the total number of sensors.

[0056] It should also be noted that dynamically adjusting sensor operating states based on risk levels and compensating for blind spots through data fusion algorithms underpin the entire intelligent security risk assessment approach. This dynamic adjustment of sensor operating states based on risk levels effectively saves energy, extends equipment life, and allows for timely adjustments to monitoring strategies based on changing risks, ensuring system reliability and security. Furthermore, compensating for blind spots through data fusion algorithms further enhances the comprehensiveness and accuracy of risk detection.

[0057] Embodiment 2 is the second embodiment of the present invention, which differs from the first two embodiments in that:

[0058] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0059] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0060] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0061] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0062] Example 3, reference Figure 2 , is an embodiment of the present invention, which provides an intelligent security risk assessment system based on multi-source data fusion, including a data processing module, a risk assessment module, and a dynamic adjustment module.

[0063] Among them, the data processing module is used to collect environmental data in real time and perform preprocessing through the deployment of sensor networks, and input the environmental data into the risk assessment model; the risk assessment module is used to extract multi-source data features through machine learning algorithms to divide risk levels and update risk assessment results in real time; the dynamic adjustment module is used to dynamically adjust the sensor operating status according to the risk level and compensate for monitoring blind spots through data fusion algorithms.

[0064] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention. It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An intelligent security risk assessment method based on multi-source data fusion, characterized in that: include: By deploying sensor networks to collect environmental data in real time and pre-processing it, the environmental data is input into the risk assessment model; Extract multi-source data features through machine learning algorithms to classify risk levels and update risk assessment results in real time; The sensor operating status is dynamically adjusted according to the risk level, and the monitoring blind spots are compensated through data fusion algorithms.

2. The intelligent security risk assessment method based on multi-source data fusion according to claim 1, characterized in that: The sensor network includes cameras, radars, infrared sensors, temperature and humidity sensors, and sound sensors.

3. The intelligent security risk assessment method based on multi-source data fusion according to claim 2, characterized in that: The collecting of environmental data and preprocessing thereof includes equipping each sensor with an independent data collection module and transmitting the collected data to a central processing unit or an edge computing node in real time via a 5G network; Ensure data transmission security through encryption protocol and compress data through LZ77 algorithm; The raw data is denoised, formatted in a unified manner, and timestamps are synchronously sampled using the NTP protocol to generate a multi-source data set with a unified time axis for input into the risk assessment model.

4. The intelligent security risk assessment method based on multi-source data fusion according to claim 3 is characterized by: The risk assessment model includes building a risk assessment model based on a deep neural network; The input layer receives multi-source data, and the hidden layer contains three fully connected layers with ReLU as the activation function; Extract features from multi-source data, and use the Softmax function to divide the risk assessment level into different levels at the output layer; Risk assessment levels include low risk c1, medium risk c2, and high risk c3; The risk level probability distribution is generated by the Softmax function, and the risk level probability is expressed as: Among them, P(c) represents the probability of the risk level of category c, the output range is [0,1], the sum of the probabilities is 1, z c represents the unnormalized model prediction value of risk level c; If P(c)>0.8, it is judged as low risk; If 0.5<P(c)≤0.8, it is judged as medium risk; If P(c)≤0.5, it is judged as high risk; The evaluation results are updated every 200ms.

5. The intelligent security risk assessment method based on multi-source data fusion according to claim 4 is characterized in that: The real-time updating of risk assessment results includes deploying a risk assessment model on an edge computing node, triggering a real-time computing thread when sensor network data is input, and setting the computing thread priority to the highest; If the edge node has insufficient computing resources, the data is forwarded to the cloud server through the message queue. The cloud server uses a distributed computing framework to process multi-node data in parallel; The assessment results are pushed to the central processing unit via the MQTT protocol and compared with the event records in the historical database. If the assessment is high risk for three consecutive times and there are similar intrusion events in the historical records, the alarm escalation mechanism is triggered and the risk level is increased by one level.

6. The intelligent security risk assessment method based on multi-source data fusion according to claim 5, characterized in that: Dynamically adjusting the operating status of sensors according to the risk level includes turning off the camera and sound sensor if the risk level is assessed as low risk, and retaining the radar, infrared sensor, and temperature and humidity sensor; If the risk level is assessed as medium, 50% of the cameras will be turned on according to the alternating polling strategy, and the sound sensor will be activated according to the threshold; If the risk level is assessed as high, all sensors are turned on, the camera switches to high frame rate mode, the radar scanning frequency is increased to 10Hz, and the sound sensor uses the directional adaptation function; When the sensor state switches, the controller checks the current energy consumption. If the total power consumption exceeds the safety threshold, non-critical sensors are forced to shut down according to priority. The sensor close and open commands are sent to the sensor controller via the Modbus protocol.

7. The intelligent security risk assessment method based on multi-source data fusion according to claim 6, characterized in that: The compensating the monitoring blind area by the data fusion algorithm includes starting the data fusion algorithm when the camera is turned off; The data fusion algorithm includes radar and infrared complementary strategy and redundancy elimination strategy; The radar and infrared complementary strategy includes the coordinates of the moving object detected by the radar (x r ,y r ) and infrared heat source coordinates (x i ,y i ) for spatial matching, if |x r -x i |≤0.5m and |y r -y i |≤0.5m, the moving objects are determined to be the same target, and the target threat index T is calculated, which is expressed as: T=αv+βΔT Where v represents the moving speed of the moving object, ΔT represents the temperature difference of the heat source, α and β represent the weight coefficients, α = 0.6, β = 0.4; The redundancy elimination strategy includes using the Hungarian algorithm to perform optimal matching on targets detected repeatedly by multiple sensors and retaining the data with the highest confidence. The confidence C is expressed as: Among them, d k represents the kth sensor measurement value, μ represents the mean, σ represents the standard deviation, and n represents the total number of sensors.

8. A system using the intelligent security risk assessment method based on multi-source data fusion according to any one of claims 1 to 7, characterized in that: Including data processing module, risk assessment module, dynamic adjustment module; The data processing module is used to collect environmental data in real time through the deployment of sensor networks and perform preprocessing, and input the environmental data into the risk assessment model; The risk assessment module is used to extract multi-source data features through machine learning algorithms to classify risk levels and update risk assessment results in real time; The dynamic adjustment module is used to dynamically adjust the sensor operating state according to the risk level and compensate for the monitoring blind area through the data fusion algorithm.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the intelligent security risk assessment method based on multi-source data fusion according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the intelligent security risk assessment method based on multi-source data fusion according to any one of claims 1 to 7 are implemented.

Images