A smart park supervision method and system based on the Internet of Things

Abstract

The present invention discloses a smart park supervision method and system based on the Internet of Things, and relates to the field of regional supervision technology. A smart park supervision method based on the Internet of Things, comprising the following steps: S1: obtaining the area to be detected and the balance processing area in the park based on the Internet of Things monitoring platform, and extracting the monitoring information in the area to be detected in real time; S2: using fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area, obtaining the required walking distance of the inspection robot in each sub-area, and using the power accumulation formula to obtain the carrying power of the inspection robot after reaching the target area to be detected according to the required walking distance of the inspection robot in each sub-area. The present invention combines the robot's static electricity accumulation, charge bearing capacity and path weight calculation by combining the weight calculation formula and the Dijkstra algorithm to generate the optimal path, which not only ensures the static electricity risk control, but also improves the inspection efficiency.

Description

Technical Field

[0001] The present invention relates to the field of regional supervision technology, and in particular to a smart park supervision method and system based on the Internet of Things. Background Art

[0002] With the development of information technology, the Internet of Things technology has gradually penetrated into all walks of life, especially in the construction and management of smart parks. The traditional park management method has many shortcomings. For example, the inspection work relies on manual labor, which is inefficient and difficult to achieve all-weather and all-round monitoring. With the advancement of smart devices and big data analysis technology, the use of smart park supervision methods based on the Internet of Things has become the key. This method can significantly improve the level of intelligence in park management.

[0003] However, there are still some challenges in implementing IoT-based smart park supervision, such as how to efficiently plan the inspection routes of inspection robots and how to protect electronic equipment in the park from electrostatic damage while ensuring inspection efficiency. Existing solutions often lack comprehensive consideration of these key issues, resulting in low inspection efficiency and certain safety hazards. Summary of the invention

[0004] In order to overcome the shortcoming of ignoring the impact of accumulated electricity on inspection robots and park electronic equipment when planning inspection routes, the present invention provides a smart park supervision method and system based on the Internet of Things.

[0005] The technical solution is: a smart park supervision method based on the Internet of Things, including the following steps:

[0006] S1: Based on the IoT monitoring platform, the detection area and the balance processing area in the park are obtained, and the monitoring information in the detection area is extracted in real time;

[0007] S2: Use fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area, and obtaining the required walking distance of the inspection robot in each sub-area. According to the required walking distance of the inspection robot in each sub-area, use the power accumulation formula to obtain the carrying power of the inspection robot after reaching the target area to be inspected;

[0008] S3: combining the inspection robot position and the carried power, using a weight calculation formula to obtain the weights of each route from the inspection robot position to the target inspection area, and using the Dijkstra algorithm to obtain the optimal auxiliary inspection route;

[0009] S4: Adjust the number of inspection robots that enter based on the corresponding capacity of each area to be inspected at different times to obtain the maximum inspection number.

[0010] Preferably, the Internet of Things monitoring platform is used to obtain the area to be detected and the balance processing area in the park, and to extract the monitoring information in the area to be detected in real time, including: using the FIFO algorithm to obtain a monitoring sequence of a preset length, and inputting the security monitoring data collected from the Internet of Things monitoring platform into each monitoring sequence, and then judging the monitoring sequence. When the ratio of the number of abnormal data in the monitoring sequence to the preset length is greater than or equal to a preset threshold, it is used as an abnormal monitoring sequence, and the area corresponding to the abnormal monitoring sequence is used as the area to be detected.

[0011] Preferably, the Internet of Things monitoring platform is based on obtaining the area to be detected and the balanced processing area in the park, and extracting the monitoring information in the area to be detected in real time, including: obtaining sensor data of each environmental area in the park, the sensor data including ambient temperature, ambient humidity, material properties and air conductivity; normalizing the sensor data, and mapping each environmental feature data to a unified numerical range; using SVM to build a processing area detection model, inputting the normalized real-time environmental features into the processing area detection model, outputting the classification result of whether the current area belongs to the balanced processing area or the unbalanced processing area, and displaying the identified balanced processing areas in the Internet of Things monitoring platform according to the classification result.

[0012] Preferably, the use of SVM to construct a processing area detection model includes: using annotated historical data and expert experience to determine a training set, and performing model training through a Gaussian kernel support vector machine; optimizing a hyperplane to distinguish between balanced processing areas and unbalanced processing areas.

[0013] Preferably, the method of using fuzzy reasoning to obtain the environmental size of each area, thereby dividing the range of each area and obtaining the required walking distance of the inspection robot in each sub-area includes:

[0014] Obtain the ambient temperature, ambient humidity, material properties and air conductivity in each area, and fuzzify the four input variables of ambient temperature, ambient humidity, material properties and air conductivity, and convert them from specific numerical forms to membership values ​​of fuzzy sets;

[0015] Formulate fuzzy rules based on expert domain experience knowledge and historical data;

[0016] Fuzzy rules are activated by fuzzy logic operation using the minimum value method, and the membership values ​​of different rule outputs are calculated;

[0017] Based on the activation rules and the historical influence range, defuzzification is performed through the centroid method, and finally the regional influence range at the current time is obtained. According to the regional influence range, the walking distance required by the inspection robot in each sub-area is obtained.

[0018] Preferably, the method of using a power accumulation formula to obtain the power carried by the inspection robot after it reaches the target area to be inspected according to the required walking distance of the inspection robot in each sub-area includes: wherein the power accumulation formula is:

[0019] ;

[0020] In the formula, For experienced The amount of charge accumulated on the inspection robot after entering each environmental area; The total number of environmental areas that the inspection robot experiences on its way to the target area; The first step that the inspection robot needs to go through environmental areas; For the The distance that the inspection robot needs to walk in each environmental area; Indicates that the inspection robot is When there are multiple environmental areas, the function of the required walking distance and the charge accumulation in the corresponding area; express The actual walking distance of the inspection robot is The corresponding function value when .

[0021] Preferably, the weight calculation formula is used to obtain the weights of each route from the inspection robot position to the target inspection area by combining the inspection robot position and the carried power, and the Dijkstra algorithm is used to obtain the optimal auxiliary inspection route, including: wherein the weight calculation formula is:

[0022] ;

[0023] In the formula, is the edge weight of the inspection robot reaching the target area; is the adjustment factor; Experienced in patrol robots The amount of charge accumulated on the inspection robot after entering each environmental area; is the standard charge tolerance of the target area; Experienced in patrol robots The maximum charge sustained during the process in each environmental area; The standard charge capacity of the inspection robot; is the distance between the inspection robot and the target area; is a preset constant.

[0024] Preferably, the method combines the inspection robot position and the carried power, uses a weight calculation formula to obtain the weights of each route in the process from the inspection robot position to the target area to be inspected, and uses the Dijkstra algorithm to obtain the optimal auxiliary inspection route, including: obtaining the inspection robot position information, initial charge, standard charge of the target area and standard charge of the inspection robot, and using the power accumulation formula to calculate the inspection robot after it has experienced The maximum charge that the inspection robot can bear during the process of moving from one environmental area to another is calculated, and the weight calculation formula is used to obtain the edge weights of each route in the process from the inspection robot position to the target area to be inspected. The Dijkstra algorithm is used to obtain the optimal auxiliary inspection route.

[0025] Preferably, the number of inspection robots entering is adjusted based on the corresponding accommodation quantity of each area to be inspected at different times to obtain the maximum inspection number, including: obtaining the actual charge amount and the regional standard charge amount in each area to be inspected in real time, taking the difference between the actual charge amount and the regional standard charge amount as the accommodation quantity, calculating the maximum number of inspection robots that can enter, and using it as the maximum inspection number, wherein the maximum inspection number needs to be rounded down.

[0026] A smart park supervision system based on the Internet of Things, comprising:

[0027] The data acquisition module is used to collect sensor data in the park, including ambient temperature, humidity, material properties, and air conductivity;

[0028] The regional node acquisition module uses the IoT monitoring platform to obtain the area to be detected and the balance processing area;

[0029] The dangerous area classification submodule uses the IoT monitoring platform to obtain the areas to be detected and the abnormal monitoring sequence;

[0030] The balanced processing area submodule uses the SVM model to distinguish the balanced processing area from the unbalanced processing area based on the sensor data;

[0031] The environment division module uses fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area;

[0032] The walking distance acquisition module is used to obtain the walking distance required by the inspection robot in each sub-area;

[0033] The power management module is used to calculate the charge accumulation of the inspection robot in multiple areas and provide data support for path optimization;

[0034] The path optimization module is used to obtain the optimal auxiliary inspection route using the Dijkstra algorithm based on the power and weight formula;

[0035] The inspection quantity dynamic adjustment module is used to adjust the number of inspection robots according to the capacity of the area to be inspected.

[0036] Beneficial effects of the present invention:

[0037] 1. The present invention uses the FIFO algorithm to process abnormal monitoring data, obtains the area to be detected, and analyzes the environmental data in combination with the SVM model, so as to accurately identify the balanced processing area that needs to be monitored. According to the area to be detected and the balanced processing area, the pertinence and effectiveness of monitoring are improved;

[0038] 2. The present invention dynamically calculates the static electricity accumulation amount of the inspection robot in different areas through the power accumulation formula, optimizes the inspection path in combination with the static electricity tolerance capacity of the area, and significantly reduces the risk of static electricity damaging the inspection robot and electronic equipment in the target area;

[0039] 3. Combining the weight calculation formula and Dijkstra algorithm, the static electricity accumulation of the robot, the standard charge tolerance of the inspection robot, the standard charge tolerance of the target area and the path weight calculation are integrated to generate the optimal path, which not only ensures the control of static electricity risks, but also improves the inspection efficiency;

[0040] 4. Dynamically adjust the number of inspection robots and allocate inspection resources according to the charge capacity of the target area, thereby improving the inspection coverage and efficiency of the park while ensuring safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of a smart park supervision method based on the Internet of Things of the present invention;

[0042] Figure 2 This is a flow chart of a smart park supervision system based on the Internet of Things of the present invention;

[0043] Figure 3 A block diagram of a regional node acquisition module of the present invention;

[0044] Figure 4 This is an example diagram of dividing the scope of each area in the present invention;

[0045] Figure 5 This is an example diagram of the walking distance required by the inspection robot in each sub-area according to the present invention. DETAILED DESCRIPTION

[0046] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0047] Embodiment 1: A method for supervising a smart park based on the Internet of Things, such as Figure 1-Figure 5 As shown, the following steps are included:

[0048] S1: Based on the IoT monitoring platform, the detection area and the balance processing area in the park are obtained, and the monitoring information in the detection area is extracted in real time;

[0049] A monitoring sequence of preset length is obtained using the FIFO algorithm. After the security monitoring data collected from the IoT monitoring platform is input into each monitoring sequence, the monitoring sequence is judged. When the ratio of the number of abnormal data in the monitoring sequence to the preset length is greater than or equal to a preset threshold, it is used as an abnormal monitoring sequence, and the area corresponding to the abnormal monitoring sequence is used as the area to be detected.

[0050] It should be noted that the security monitoring data, such as ambient temperature, ambient humidity, equipment operating status, and electrical parameters, are obtained through the Internet of Things monitoring platform. Temperature and humidity sensors are used to evenly distribute the sensors in the monitored area to ensure that the uniform changes in temperature and humidity in the area can be captured, thereby obtaining the ambient temperature and humidity. The relevant parameters are obtained through the devices connected to the Internet of Things, and these data are input into a fixed-length FIFO monitoring sequence in chronological order. For example, assuming that the monitoring sequence length is 100, the latest data will overwrite the earliest data, and the sequence length is always kept fixed. The data in the monitoring sequence is analyzed one by one, and abnormal data (such as values ​​beyond the safety range) are marked. The ratio of the number of abnormal data to the total length of the sequence is calculated. If the ratio is greater than or equal to the set threshold (such as 20%), the sequence is considered to be an abnormal monitoring sequence, and the area corresponding to the abnormal monitoring sequence will be marked as an area to be detected, and the inspection robot or the park management system will be notified as a key monitoring object.

[0051] The sensor data of each environmental area in the park is obtained, and the sensor data includes environmental temperature, environmental humidity, material properties and air conductivity; the sensor data is normalized to map each environmental feature data to a unified numerical range; a processing area detection model is constructed using SVM, and the normalized real-time environmental features are input into the processing area detection model, and the classification result of whether the current area belongs to a balanced processing area or an unbalanced processing area is output, and the identified balanced processing areas are displayed in the Internet of Things monitoring platform according to the classification result.

[0052] It should be noted that the balanced processing area is the area where the amount of charge on the inspection robot is reduced. Environmental sensor data is obtained, and material conductivity sensors (such as four-probe testers or surface resistance testers) are used. The sensors are installed in fixed areas and the material properties of the target objects are scanned regularly. The data is mainly used to determine the conductive properties of the materials in the area, predict static electricity risks, and obtain material properties. Air conductivity sensors (such as thin film ion conductivity testers) are used and installed at key locations in the area, especially in environments with low humidity or dry air. The sensors capture the relationship between ion concentration and conductivity in the air and calculate the conductivity of the air in real time to obtain the air conductivity. The collected data of different physical quantities are normalized, and the annotated historical data are divided into training sets and test sets at a ratio of 8:2 to ensure that each type of area (balanced processing and unbalanced processing areas) has a sufficient number of samples to reduce category imbalance. The real-time collected regional data is input into the trained SVM model, and the classification results are output. The test data set is used to calculate the accuracy, recall rate, and F1 score to evaluate the model performance. If the accuracy is low, readjust the hyperparameters or increase the number of samples.

[0053] The training set is determined by using labeled historical data and expert experience, and the model is trained using a Gaussian kernel support vector machine. The hyperplane is optimized to distinguish between balanced processing areas and unbalanced processing areas.

[0054] It should be noted that the Gaussian kernel function (RBF kernel) is used to construct the SVM classifier to determine the optimal hyperplane, distinguish the balanced processing area from the unbalanced processing area, optimize the kernel function parameters (such as C value and γ value), and improve the classification accuracy.

[0055] S2: Use fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area, and obtaining the required walking distance of the inspection robot in each sub-area. According to the required walking distance of the inspection robot in each sub-area, use the power accumulation formula to obtain the carrying power of the inspection robot after reaching the target area to be inspected;

[0056] Obtain the ambient temperature, ambient humidity, material properties and air conductivity in each area, and fuzzify the four input variables of ambient temperature, ambient humidity, material properties and air conductivity, and convert them from specific numerical forms to membership values ​​of fuzzy sets;

[0057] Formulate fuzzy rules based on expert domain experience knowledge and historical data;

[0058] Fuzzy rules are activated by fuzzy logic operation using the minimum value method, and the membership values ​​of different rule outputs are calculated;

[0059] Based on the activation rules and the historical influence range, defuzzification is performed through the centroid method, and finally the regional influence range at the current time is obtained. According to the regional influence range, the walking distance required by the inspection robot in each sub-area is obtained.

[0060] It should be noted that the environmental characteristics (temperature, humidity, material properties, air conductivity) are used as input variables, these variables are fuzzified through fuzzy logic methods, fuzzy rules are formulated, the regional influence range is derived by combining the historical influence range and current data, and the walking distance required by the inspection robot is calculated based on the regional influence range; a specific example is to use IoT sensors to collect environmental temperature, humidity, material properties and air conductivity data in different areas of the park; the environmental characteristic variables are converted into fuzzy sets, and three membership values ​​of low, medium and high are defined. For example: fuzzy membership function of temperature: "Low temperature": applicable to the range of 15°C-25°C, expressed by trigonometric functions. "Medium temperature": applicable to the range of 20°C-30°C, expressed by trapezoidal functions. "High temperature": Applicable to the range of 25°C-35°C. The same is true for other variables. Use expert experience and historical data to define appropriate membership functions. Formulate fuzzy rules based on expert knowledge and historical data. For example, if the humidity is high and the material properties are moderate, the regional impact range is large; if the air conductivity is low and the temperature is moderate, the regional impact range is moderate. Use the minimum method to activate the fuzzy rules and calculate the membership values ​​of different rule outputs. For example, if the humidity membership value is 0.8 and the material property membership value is 0.6, then the output membership of the rule with high humidity and moderate material properties is min (0.8, 0.6) = 0.6. Use the centroid method to comprehensively calculate the membership values ​​of the activated rules to determine the regional impact range at the current time. For example, if the output membership values ​​of the activated rules are 0.6, 0.4, and 0.2, respectively, the regional impact range is calculated as 15m by weighted average.

[0061] The power accumulation formula is:

[0062] ;

[0063] In the formula, For experienced The amount of charge accumulated on the inspection robot after entering each environmental area; The total number of environmental areas that the inspection robot experiences on its way to the target area; The first step that the inspection robot needs to go through environmental areas; For the The distance that the inspection robot needs to walk in each environmental area; Indicates that the inspection robot is When there are multiple environmental areas, the function of the required walking distance and the charge accumulation in the corresponding area; express The actual walking distance of the inspection robot is The corresponding function value when .

[0064] It should be noted that the target area is the area to be detected or the balanced processing area. The total number of areas that the inspection robot should go through from its position to reach the area to be detected or the balanced processing area is , by calculating the total charge accumulated in the detection area and the charge eliminated in the equilibrium processing area, For experienced The amount of charge accumulated on the inspection robot after each environmental area, and the functions used in different areas to be inspected or balanced processing areas should be their respective related functions.

[0065] S3: combining the inspection robot position and the carried power, using a weight calculation formula to obtain the weights of each route from the inspection robot position to the target inspection area, and using the Dijkstra algorithm to obtain the optimal auxiliary inspection route;

[0066] The weight calculation formula is:

[0067] ;

[0068] In the formula, is the edge weight of the inspection robot reaching the target area; is the adjustment factor; Experienced in patrol robots The amount of charge accumulated on the inspection robot after entering each environmental area; is the standard charge tolerance of the target area; Experienced in patrol robots The maximum charge sustained during the process in each environmental area; The standard charge capacity of the inspection robot; is the distance between the inspection robot and the target area; is a preset constant.

[0069] Obtain the inspection robot's position information, initial charge, standard charge capacity of the target area, and standard charge capacity of the inspection robot, and use the power accumulation formula to calculate the inspection robot's The maximum charge that the inspection robot can bear during the process of moving from one environmental area to another is calculated, and the weight calculation formula is used to obtain the edge weights of each route in the process from the inspection robot position to the target area to be inspected. The Dijkstra algorithm is used to obtain the optimal auxiliary inspection route.

[0070] It should be noted that the formula is not only used to evaluate the deviation between the current charge of the robot and the standard charge of the target area. The closer to the standard value, the lower the weight, indicating a better path. It is also used to evaluate whether the maximum charge borne by the robot on the path exceeds its upper limit, to avoid the risk of static electricity caused by path selection, to consider the path distance, and to weigh the priority of the shortest path. The edge weights of all paths from the current position of the inspection robot to the target area are calculated to generate a weight graph, and the Dijkstra algorithm is used to calculate the path with the smallest total weight from the weight graph as the optimal auxiliary inspection route.

[0071] S4: Adjust the number of inspection robots that enter based on the corresponding capacity of each area to be inspected at different times to obtain the maximum inspection number.

[0072] The actual charge amount and the regional standard charge amount in each area to be inspected are obtained in real time, and the difference between the actual charge amount and the regional standard charge amount is used as the accommodation quantity. The maximum number of inspection robots that can enter is calculated and used as the maximum inspection quantity, where the maximum inspection quantity needs to be rounded down.

[0073] It should be noted that the actual charge is the total charge accumulated due to environmental factors or other charges carried by the inspection robots that have entered the real-time monitoring area. The data source is obtained through the Internet of Things platform in real time by the electrostatic sensors, environmental temperature and humidity sensors and air conductivity detection devices installed in the area. The regional standard charge is the maximum safe charge that the area can withstand, which is pre-set to ensure that electronic equipment or components in the area are not damaged by excessive static electricity. The charge capacity of the area is calculated, and the capacity is defined as the difference between the regional standard charge and the actual charge: ; In the formula, To accommodate the quantity, is the standard charge of the region, is the actual charge amount; by rounding down, it is ensured that the actual number of robots entering does not exceed the upper limit of the area's charge capacity and the minimum number of robots entering should be greater than or equal to 1.

[0074] Embodiment 2: Based on Embodiment 1, a smart park supervision system based on the Internet of Things includes:

[0075] The data acquisition module is used to collect sensor data in the park, including ambient temperature, humidity, material properties, and air conductivity;

[0076] The regional node acquisition module uses the IoT monitoring platform to obtain the area to be detected and the balance processing area;

[0077] The dangerous area classification submodule uses the IoT monitoring platform to obtain the areas to be detected and the abnormal monitoring sequence;

[0078] The balanced processing area submodule uses the SVM model to distinguish the balanced processing area from the unbalanced processing area based on the sensor data;

[0079] The environment division module uses fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area;

[0080] The walking distance acquisition module is used to obtain the walking distance required by the inspection robot in each sub-area;

[0081] The power management module is used to calculate the charge accumulation of the inspection robot in multiple areas and provide data support for path optimization;

[0082] The path optimization module is used to obtain the optimal auxiliary inspection route using the Dijkstra algorithm based on the power and weight formula;

[0083] The inspection quantity dynamic adjustment module is used to adjust the number of inspection robots according to the capacity of the area to be inspected.

[0084] The above embodiments are provided for persons familiar with the art to implement or use the present invention. Personnel familiar with the art can make various modifications or changes to the above embodiments without departing from the inventive concept of the present invention. Therefore, the protection scope of the present invention is not limited to the above embodiments, but should be the maximum scope of the innovative features mentioned in the claims.

Claims

1. A smart park supervision method based on the Internet of Things, characterized in that: The following steps are involved: S1: Based on the IoT monitoring platform, the detection area and the balance processing area in the park are obtained, and the monitoring information in the detection area is extracted in real time; S2: Use fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area, and obtaining the required walking distance of the inspection robot in each sub-area. According to the required walking distance of the inspection robot in each sub-area, use the power accumulation formula to obtain the carrying power of the inspection robot after reaching the target area to be inspected; S3: combining the inspection robot position and the carried power, using a weight calculation formula to obtain the weights of each route from the inspection robot position to the target inspection area, and using the Dijkstra algorithm to obtain the optimal auxiliary inspection route; S4: adjusting the number of inspection robots entering based on the corresponding capacity of each area to be inspected at different times to obtain the maximum inspection number; The method of obtaining the to-be-detected area and the balance processing area in the park based on the Internet of Things monitoring platform and extracting the monitoring information in the to-be-detected area in real time includes: using a FIFO algorithm to obtain a monitoring sequence of a preset length, inputting the security monitoring data collected from the Internet of Things monitoring platform into each monitoring sequence, and then judging the monitoring sequence. When the ratio of the number of abnormal data in the monitoring sequence to the preset length is greater than or equal to a preset threshold, the monitoring sequence is regarded as an abnormal monitoring sequence, and the area corresponding to the abnormal monitoring sequence is regarded as the to-be-detected area; According to the required walking distance of the inspection robot in each sub-area, the power accumulation formula is used to obtain the power carried by the inspection robot after reaching the target area to be inspected, wherein the power accumulation formula is: ； In the formula, For experienced The amount of charge accumulated on the inspection robot after entering each environmental area; The total number of environmental areas that the inspection robot experiences on its way to the target area; The first step that the inspection robot needs to go through environmental areas; For the The distance that the inspection robot needs to walk in each environmental area; Indicates that the inspection robot is When there are multiple environmental areas, the function of the required walking distance and the charge accumulation in the corresponding area; express The actual walking distance of the inspection robot is The corresponding function value when ; The weight calculation formula is used to obtain the weights of each route from the inspection robot position to the target inspection area in combination with the inspection robot position and the carried power, and the Dijkstra algorithm is used to obtain the optimal auxiliary inspection route, including: wherein the weight calculation formula is: ； In the formula, is the edge weight of the inspection robot reaching the target area; is the adjustment factor; Experienced in inspection robots The amount of charge accumulated on the inspection robot after entering each environmental area; is the standard charge tolerance of the target area; Experienced in inspection robots The maximum charge sustained during the process in each environmental area; The standard charge capacity of the inspection robot; is the distance between the inspection robot and the target area; is a preset constant; The method combines the inspection robot position and the carried power, uses a weight calculation formula to obtain the weights of each route in the process from the inspection robot position to the target area to be inspected, and uses the Dijkstra algorithm to obtain the optimal auxiliary inspection route, including: obtaining the inspection robot's position information, initial charge, standard charge capacity of the target area and standard charge capacity of the inspection robot, and using the power accumulation formula to calculate the inspection robot after it has experienced The maximum charge that the inspection robot can bear during the process of reaching the target inspection area is calculated by the weight calculation formula, and the edge weight of each route is obtained by using the Dijkstra algorithm. The balancing treatment area is an area where the amount of electric charge on the inspection robot is reduced.

2. According to the method for monitoring a smart park based on the Internet of Things according to claim 1, it is characterized in that: The method of obtaining the to-be-detected area and the balanced processing area in the park based on the Internet of Things monitoring platform, and extracting the monitoring information in the to-be-detected area in real time, includes: obtaining sensor data of each environmental area in the park, wherein the sensor data includes environmental temperature, environmental humidity, material properties and air conductivity; normalizing the sensor data, and mapping each environmental feature data to a unified numerical range; using SVM to build a processing area detection model, inputting the normalized real-time environmental features into the processing area detection model, outputting the classification result of whether the current area belongs to the balanced processing area or the unbalanced processing area, and displaying the identified balanced processing areas in the Internet of Things monitoring platform according to the classification result.

3. According to the method for monitoring a smart park based on the Internet of Things according to claim 2, it is characterized in that: The method of using SVM to construct a processing area detection model includes: using annotated historical data and expert experience to determine a training set, and performing model training through a Gaussian kernel support vector machine; and optimizing a hyperplane to distinguish between a balanced processing area and an unbalanced processing area.

4. According to the method for monitoring a smart park based on the Internet of Things according to claim 1, it is characterized in that: The use of fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area and obtaining the required walking distance of the inspection robot in each sub-area, includes: Obtain the ambient temperature, ambient humidity, material properties and air conductivity in each area, and fuzzify the four input variables of ambient temperature, ambient humidity, material properties and air conductivity, and convert them from specific numerical forms to membership values ​​of fuzzy sets; Formulate fuzzy rules based on expert domain experience knowledge and historical data; Fuzzy rules are activated by fuzzy logic operation using the minimum value method, and the membership values ​​of different rule outputs are calculated; Based on the activation rules and the historical influence range, defuzzification is performed through the centroid method, and finally the regional influence range at the current time is obtained. According to the regional influence range, the walking distance required by the inspection robot in each sub-area is obtained.

5. According to the method for monitoring a smart park based on the Internet of Things according to claim 1, it is characterized in that: The method adjusts the number of inspection robots entering based on the corresponding accommodation number of each area to be inspected at different times to obtain the maximum inspection number, including: obtaining the actual charge amount and the regional standard charge amount in real time in each area to be inspected, taking the difference between the actual charge amount and the regional standard charge amount as the accommodation number, calculating the maximum number of inspection robots that can enter, and using it as the maximum inspection number, wherein the maximum inspection number needs to be rounded down.

6. A smart park supervision system based on the Internet of Things, according to a smart park supervision method based on the Internet of Things according to any one of claims 1-5, characterized in that: include: The data acquisition module is used to collect sensor data in the park, including ambient temperature, humidity, material properties, and air conductivity; The regional node acquisition module uses the IoT monitoring platform to obtain the area to be detected and the balanced processing area. The area corresponding to the abnormal monitoring sequence will be marked as the area to be detected, and the balanced processing area is the area where the charge on the inspection robot is reduced; The dangerous area classification submodule uses the IoT monitoring platform to obtain the areas to be detected and the abnormal monitoring sequence; The balanced processing area submodule uses the SVM model to distinguish the balanced processing area from the unbalanced processing area based on the sensor data; The environment division module uses fuzzy reasoning to obtain the environmental size of each area, thereby dividing the scope of each area; The walking distance acquisition module is used to obtain the walking distance required by the inspection robot in each sub-area; The power management module is used to calculate the charge accumulation of the inspection robot in multiple areas and provide data support for path optimization; The path optimization module is used to obtain the optimal auxiliary inspection route using the Dijkstra algorithm based on the power and weight formula; The inspection quantity dynamic adjustment module is used to adjust the number of inspection robots according to the capacity of the area to be inspected.

Images