Escalator and moving sidewalk environment monitoring system based on Internet of Things technology

Abstract

The invention discloses an escalator and moving sidewalk environment monitoring system based on the Internet of Things technology, and relates to the technical field of escalator environment parameter monitoring, and the escalator and moving sidewalk environment monitoring system comprises the following steps: obtaining environment monitoring data collected by sensors distributed on equipment; establishing a sensor network, remotely transmitting environmental monitoring data collected by each sensor in a target area by combining a general packet radio service technology, and storing the environmental monitoring data in a database of a server; performing real-time analysis on the environment monitoring data acquired by each sensor by using a preset model, and determining an early warning stage of the environment monitoring data; and matching the target early-warning mode from a preset early-warning mode library to carry out early-warning operation, and carrying out synchronous early-warning notification on the management terminal. According to the invention, real-time monitoring, efficient data transmission, intelligent analysis and early warning of the environmental data of the escalator and the moving sidewalk are realized, the safety of a user when the escalator and the moving sidewalk are used is improved, and the possibility of accidents is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of escalator environmental parameter monitoring, and more specifically to an escalator and moving walkway environmental monitoring system based on Internet of Things technology. Background Art

[0002] Escalators are widely used in various occasions, such as department stores, office buildings and public facilities, as well as various indoor or outdoor environments. The safe and stable operation of escalators requires monitoring the operating status of relevant components, and the operating status of relevant components is closely related to the application occasions of escalators, the temperature, humidity and dust in the environment, and is also closely related to the number of users and the usage frequency. Therefore, the status of each escalator is different, and it is necessary to frequently conduct on-site inspections and maintenance of relevant components to ensure the correct operation of relevant components. This involves a large amount of work, and it is difficult for escalator manufacturers or users to keep track of their operating status in an organized manner.

[0003] Therefore, how to provide an escalator and moving walkway environmental monitoring system based on Internet of Things technology to realize real-time monitoring, efficient data transmission, intelligent analysis and early warning of the environmental data of escalators and moving walkways, improve the safety of users when using escalators and moving walkways, and reduce the possibility of accidents is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0004] In view of this, the present invention provides an escalator and moving walkway environmental monitoring system based on Internet of Things technology, which monitors the operating environment of escalators and moving walkways, and uses infrared wireless sensing technology to monitor temperature, humidity, rainfall, and whether there are passengers through infrared monitoring. In harsh environments, warning notifications are sent through the Internet of Things system of escalators and moving walkways to the user unit and the maintenance unit to manually stop the escalator or take preventive measures; or when no passengers are identified, the escalator is intelligently controlled to stop through the Internet of Things system. At the same time, the obtained monitoring data provides a basis for the life research of the main components and safety protection devices of escalators in different environments.

[0005] To achieve the above object, the present invention adopts the following technical solutions: An escalator and moving walkway environmental monitoring system based on Internet of Things technology, including:

[0006] An environmental monitoring module, configured to obtain environmental monitoring data collected by each sensor distributed on the device at a preset frequency;

[0007] A data transmission module, configured to form a sensor network, combine General Packet Radio Service (GPRS) technology, remotely transmit the environmental monitoring data collected by each sensor in the target area, and store it in the database of the server;

[0008] A data analysis module, which is used to perform real-time analysis on the environmental monitoring data collected by each sensor by using a preset model, and determine the early warning stage of the environmental monitoring data when the state change trend of the data meets the early warning conditions;

[0009] An early warning module, which is used to determine the early warning requirements at the early warning stage, match the target early warning method from the preset early warning method library for early warning operations, and synchronously send early warning notifications to the management terminal.

[0010] Preferably, it further includes: a visualization module, which is used to display the operation status of the device and the data analysis results in real time, and display early warning notifications. It is convenient for management personnel to monitor the device status and environmental data of the escalator in real time, and is convenient for adjusting management strategies at any time.

[0011] Preferably, the sensors include a temperature sensor, a humidity sensor, a rainfall sensor and an infrared sensor;

[0012] The temperature sensor monitors the ambient temperature of the device operation in real time;

[0013] The humidity sensor monitors the air humidity of the device operation in real time;

[0014] The rainfall sensor monitors the ambient rainfall of the device operation in real time;

[0015] The infrared sensor monitors the operation of passengers on the device in real time.

[0016] Preferably, the data analysis module includes: an environment analysis sub-module, a personnel monitoring sub-module, an anomaly analysis sub-module and a decision-making sub-module;

[0017] The environment analysis sub-module is used to receive the ambient temperature, air humidity and ambient rainfall data, and generate a temperature change trend graph, a humidity change trend graph and a rainfall change trend graph respectively;

[0018] The personnel monitoring sub-module is used to receive the operation of passengers, analyze the peak hours and crowd density of personnel, obtain a personnel flow trend graph, and control the device to shut down when no passengers are detected, and control the device to operate normally when passengers are identified;

[0019] The anomaly analysis sub-module is used to aggregate the ambient temperature, humidity and rainfall data as inputs, train a machine learning model for prediction, and apply an anomaly detection algorithm to identify abnormal patterns in the data;

[0020] The decision-making sub-module is used to determine the device operation strategy by combining the results of environment analysis, personnel monitoring and anomaly analysis.

[0021] Preferably, aggregate environmental temperature, humidity, and rainfall data as input, train a machine learning model for prediction, and apply an anomaly detection algorithm to identify abnormal patterns in the data, including:

[0022] Obtain a training sample set, a test sample set, and a feature set;

[0023] Use the decision tree algorithm to train the training sample set, construct a decision tree, and the construction process includes the adaptive threshold of the feature attributes of the training sample set, the selection of the feature attributes of the decision tree node, and the recursive formation of the decision tree based on the feature attribute threshold and the node feature attributes;

[0024] Use the test sample set to test the decision tree, check whether the result meets the preset standard, if it meets, use the decision tree to predict the environmental data;

[0025] If the test result does not meet the preset standard, add training samples to the training sample set, and re-judge through the above steps until the test result meets the preset standard. Timely adding new training samples to the training sample set is beneficial to adjusting a decision tree that meets the actual classification standard. The new training samples are obtained from the environmental monitoring module;

[0026] Prune the decision tree to obtain an anomaly recognition model.

[0027] Preferably, the adaptive threshold of the feature attributes includes: for each feature, calculate its splitting threshold by traversing the feature values and calculating each split point; use information gain, Gini index, or other evaluation metrics to evaluate the splitting effect of each feature.

[0028] The selection of the feature attributes of the decision tree node includes: according to the calculated feature attribute threshold, select the best feature as the splitting feature of the current node; the best feature is the feature with the largest information gain or the smallest Gini index; divide the sample set according to the selected feature attribute and threshold to form subsets;

[0029] The recursive formation of the decision tree includes: for each subset, check whether the stopping condition is met (such as the purity of the sample set reaches a certain degree, the features are exhausted, etc.); if the stopping condition is met, mark the current node as a leaf node and assign the corresponding class label; if the stopping condition is not met, repeat the above operations to continue feature selection and division for each subset until the complete decision tree is constructed.

[0030] Preferably, it further includes: optimizing the decision tree using the grid search method, inputting the monitored environmental data into the optimized decision tree, and obtaining optimized environmental information for environmental anomaly recognition.

[0031] Preferably, pruning the decision tree to obtain an anomaly recognition model includes:

[0032] There are two pruning methods: non - pruning and pruning; there are two pruning algorithms. The error - reduction algorithm prunes according to the number of clusters. If the number of clusters is set to 3, randomly select 2 samples for training and use the remaining 1 sample to correct the model. The setting of the pruning cluster number is related to the amount of modeling data. If the data volume is greater than 100, set it to an integer between 2 and 100; otherwise, set it to an integer between 2 and the data volume.

[0033] Pruning by confidence boundary calculates the error estimate value of the node according to the confidence boundary, then performs pruning, and selects the node with the smallest error rate. The confidence boundary is set to a number between (0, 1).

[0034] Among them, the method for judging whether pruning is required is as follows:

[0035] Use the training data set to estimate the error before and after pruning, so as to decide whether to perform pruning:

[0036] Calculate the upper limit of the confidence level:

[0037]

[0038] Among them, N is the number of power data, Pr is the probability density function, E is the number of errors in the power consumption category attributes of N power data, q is the true error rate, c is the confidence level, the default value of c is 0.25, and z is the standard deviation corresponding to the confidence level c.

[0039] Preferably, the sensor network is located within the sensing area of the device and communicates with the environmental monitoring module through a communication master station. The sensor network is a ZigBee wireless ad - hoc network, including: sensors and a wireless gateway. The sensors are used to collect environmental data of escalators and moving walks; the wireless gateway communicates with the sensors wirelessly and communicates with the communication master station through a fieldbus. The wireless gateway transmits the environmental data to the communication master station and transmits the monitoring instructions to the sensors. Among them, the routing algorithm of the sensor network adopts the CRAM algorithm, and the sensors and the wireless gateway adopt a burst - wake - up communication mechanism, and only perform wireless network communication when the data changes abnormally within a certain time interval or according to the instructions of the environmental monitoring module.

[0040] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses an escalator and moving walkway environment monitoring system based on Internet of Things technology, including: an environment monitoring module for obtaining environment monitoring data collected by each sensor distributed on the equipment at a preset frequency; a data transmission module for forming a sensor network, combining General Packet Radio Service (GPRS) technology to remotely transmit the environment monitoring data collected by each sensor in the target area and store it in the database of the server; a data analysis module for using a preset model to perform real-time analysis on the environment monitoring data collected by each sensor, and determining the warning stage of the environment monitoring data when the state change trend of the data meets the warning conditions; a warning module for determining the warning requirements in the warning stage, matching a target warning method from a preset warning method library to perform a warning operation, and synchronously sending a warning notification to the management terminal. The present invention realizes real-time monitoring, efficient data transmission, intelligent analysis and warning of the environment data of escalators and moving walkways, improves the safety of users when using escalators and moving walkways, and reduces the possibility of accidents. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on the provided drawings.

[0042] Figure 1 It is a schematic structural diagram of the escalator and moving walkway environment monitoring system based on Internet of Things technology provided by the present invention.

[0043] Figure 2 It is a schematic structural diagram of the data analysis module provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0045] The embodiment of the present invention discloses an escalator and moving walkway environmental monitoring system based on Internet of Things technology. Through multiple sensors distributed on the equipment, various environmental factors (such as temperature, humidity, rainfall, etc.) in the target area are comprehensively monitored. The transmitted data is stored in the database of the server, which is convenient for centralized management and subsequent analysis, and improves the data processing efficiency. When the data change meets the warning conditions, the system timely determines the warning stage to effectively prevent potential safety risks. While receiving the warning notice, the management terminal can quickly take corresponding measures to reduce the occurrence probability of environmental accidents. It improves the safety of users when using escalators and moving walkways and reduces the possibility of accidents.

[0046] In a specific embodiment of the present invention, the escalator and moving walkway environmental monitoring system based on Internet of Things technology, as Figure 1 shown, includes:

[0047] An environmental monitoring module, configured to obtain the environmental monitoring data collected by each sensor distributed on the equipment at a preset frequency;

[0048] A data transmission module, configured to form a sensor network, combine with General Packet Radio Service (GPRS) technology, remotely transmit the environmental monitoring data collected by each sensor in the target area, and store it in the database of the server;

[0049] A data analysis module, configured to perform real-time analysis on the environmental monitoring data collected by each sensor using a preset model, determine the state change trend of the environmental monitoring data, and determine the warning stage of the environmental monitoring data when the state change trend of the data meets the warning conditions;

[0050] A warning module, configured to determine the warning requirements at the warning stage, match the target warning method from the preset warning method library based on the warning requirements for warning operations, and synchronously send a warning notice to the management terminal.

[0051] Specifically, it further includes: a visualization module, configured to display the running state of the equipment and the data analysis result in real time, and display the warning notice. It is convenient for management personnel to monitor the equipment state and environmental data of the escalator in real time and adjust the management strategy at any time.

[0052] Specifically, the sensors include a temperature sensor, a humidity sensor, a rainfall sensor, and an infrared sensor;

[0053] The temperature sensor monitors the environmental temperature of the equipment operation in real time; to prevent equipment overheating and resulting in failures in high-temperature situations;

[0054] The humidity sensor monitors the air humidity of the equipment operation in real time; to avoid slips and equipment failures caused by excessive moisture;

[0055] The rainfall sensor monitors the environmental rainfall in real time during the operation of the device to determine whether corresponding protective measures need to be taken.

[0056] The infrared sensor monitors the operation of passengers on the device in real time. By identifying whether there are passengers, it intelligently controls the opening and closing of the device operation. When passengers are identified, the device is controlled to start; when no passengers are identified, the device is controlled to shut down.

[0057] Specifically, the data analysis module includes: an environmental analysis sub-module, a personnel monitoring sub-module, an anomaly analysis sub-module, and a decision-making sub-module; as Figure 2 shown;

[0058] The environmental analysis sub-module is used to receive data on environmental temperature, air humidity, and environmental rainfall, and generate a temperature change trend graph, a humidity change trend graph, and a rainfall change trend graph respectively.

[0059] According to the data of the temperature sensor obtained in real time, a temperature change trend graph is generated to evaluate the environmental comfort level. The data of the humidity sensor is monitored to analyze the impact of humidity on the environment; rainfall data is collected through the rainfall sensor to analyze the impact of rainfall on the flow of people and the operation of the device. Through the temperature, humidity, and rainfall change trend graphs, real-time environmental monitoring information is provided to assist in decision-making.

[0060] The personnel monitoring sub-module is used to receive the operation of passengers, analyze the peak hours and the density of the flow of people, obtain a personnel flow trend graph, and control the device to shut down when no passengers are monitored. When passengers are identified, the device operates normally.

[0061] The personnel monitoring sub-module monitors the number and flow of passengers on the escalator in real time. According to the real-time data analysis, the peak hours and the density of the flow of people are identified to generate a personnel flow trend graph. When no passengers are monitored, the device is automatically controlled to shut down; when passengers are identified, the normal operation of the device is restored.

[0062] The anomaly analysis sub-module is used to aggregate the data of environmental temperature, humidity, and rainfall as inputs, train a machine learning model for prediction, and apply an anomaly detection algorithm to identify abnormal patterns in the data.

[0063] The decision-making sub-module is used to determine the device operation strategy by combining the results of environmental analysis, personnel monitoring, and anomaly analysis.

[0064] By effectively integrating sub-modules such as environmental analysis, personnel monitoring, anomaly analysis, and decision-making, making full use of environmental and personnel data, the intelligent monitoring and management of escalators and sidewalks are realized to ensure the safety of passengers and the efficient operation of the device.

[0065] Specifically, aggregate environmental temperature, humidity, and rainfall data as inputs, train a machine learning model for prediction, and apply an anomaly detection algorithm to identify abnormal patterns in the data, including:

[0066] Obtain a training sample set, a test sample set, and a feature set; the feature set at least includes environmental temperature: current temperature (°C or °F), historical temperature (temperature records for the previous hour, day, and week); relative humidity: current relative humidity (%), historical humidity (humidity records for the previous hour, day, and week); rainfall: current rainfall (mm), historical rainfall (rainfall records for the previous hour, day, and week); time features: specific time, day of the week, season; meteorological conditions: wind speed (m / s), air pressure (hPa); pedestrian flow data: current pedestrian flow (real-time passenger number), historical pedestrian flow (pedestrian flow data for the previous hour, day, and week); equipment status: escalator / walkway operating status (in operation, stopped, under maintenance), fault records (historical fault times and frequencies).

[0067] Use the decision tree algorithm to train the training sample set, construct a decision tree, and the construction process includes adaptive threshold of feature attributes of the training sample set, selection of decision tree node feature attributes, and recursive formation of the decision tree based on the feature attribute threshold and node feature attributes;

[0068] Use the test sample set to test the decision tree, check whether the result meets the preset standard, if it meets, use the decision tree to predict environmental data;

[0069] If the test result does not meet the preset standard, add training samples to the training sample set, and re-judge through the above steps until the test result meets the preset standard. Timely adding new training samples to the training sample set is beneficial to adjusting a decision tree that meets the actual classification standard. The new training samples are obtained from the environmental monitoring module;

[0070] Prune the decision tree to obtain an anomaly recognition model.

[0071] Specifically, the adaptive threshold of feature attributes includes: for each feature, calculate its split threshold by traversing the feature values and calculating each split point; use information gain, Gini index, or other evaluation metrics to evaluate the split effect of each feature.

[0072] The selection of decision tree node feature attributes includes: according to the calculated feature attribute threshold, select the best feature as the splitting feature of the current node; the best feature is the feature with the largest information gain or the smallest Gini index; divide the sample set according to the selected feature attribute and threshold to form subsets;

[0073] Recursively form a decision tree, including: for each subset, check whether the stopping condition is met (such as the sample set purity reaching a certain level, all features being used up, etc.); if the stopping condition is met, mark the current node as a leaf node and assign the corresponding class label; if the stopping condition is not met, repeat the above operations to continue feature selection and partitioning for each subset until the complete decision tree is constructed.

[0074] Specifically, it further includes: optimizing the decision tree using the grid search method, inputting the monitored environmental data into the optimized decision tree to obtain optimized environmental information for environmental anomaly identification. The decision tree algorithm can effectively process and analyze environmental data to identify patterns, predict trends, and detect anomalies.

[0075] Calculate the measure index of impurity;

[0076]

[0077] In the formula, Entropy represents information entropy, Gini represents the Gini index; t represents a given node, i represents any classification of the label, c represents the number of classification categories, and p(i|t) represents the proportion of i at node t;

[0078] Set the maximum depth of the decision tree to k, the random mode parameter in the decision tree branches to random, the minimum number of samples in each decision tree leaf node to min_leaf, the minimum number of samples in each non-leaf node to min_split, and the number of features considered when restricting branches to num, and set the parameter value range and step size;

[0079] Form a grid space by crossing different parameter values. The grid nodes of the grid space are the corresponding parameter pairs;

[0080] Construct a decision tree for each group of the parameter pairs and evaluate the classification effect using the test set;

[0081] Select the parameter pair with the best classification effect. If the step size meets the requirements, output the optimal parameter pair; if the step size does not meet the requirements, re-set the step size and re-form the grid space;

[0082] Set the optimal parameter pair as the parameters of the decision tree and output the optimal decision tree.

[0083] Specifically, pruning the decision tree to obtain an anomaly recognition model includes:

[0084] There are two pruning methods: non-pruning and pruning. There are two pruning algorithms. The error reduction algorithm prunes according to the number of clusters. If the number of clusters is set to 3, randomly select 2 for training and use the remaining one to correct the model. The setting of the pruning cluster number is related to the amount of modeling data. If the data volume is greater than 100, set it to an integer between 2 and 100; otherwise, set it to an integer between 2 and the data volume.

[0085] Pruning by confidence bounds calculates the error estimate of a node based on the confidence bounds and then prunes, selecting the node with the smallest error rate. The confidence bounds are set to a number between (0, 1).

[0086] Among them, the method for determining whether pruning is required is as follows:

[0087] Use the training dataset to estimate the error before and after pruning to determine whether to perform pruning:

[0088] Calculate the upper limit of the confidence level:

[0089]

[0090] Among them, N is the number of power data, Pr is the probability density function, E is the number of errors in the electricity consumption category attributes of N power data, q is the true error rate, c is the confidence level, the default value of c is 0.25, and z is the standard deviation corresponding to the confidence level c.

[0091] Specifically, the environmental monitoring module is located on the ground and installs system monitoring software and a background database. The environmental monitoring module is used to receive environmental data from sensors and issue monitoring instructions.

[0092] Specifically, it further includes: a communication master station, which communicates with the environmental monitoring module through an industrial Ethernet network and is used to receive and transmit the environmental data and the monitoring instructions.

[0093] Specifically, the sensor network is located within the sensing area of the device and communicates with the environmental monitoring module through the communication master station. The sensor network is a ZigBee wireless ad hoc network, including: sensors and a wireless gateway. The sensors are used to collect environmental data of escalators and moving walks. The wireless gateway communicates with the sensors wirelessly and communicates with the communication master station through a fieldbus. The wireless gateway transmits the environmental data to the communication master station and transmits the monitoring instructions to the sensors. Among them, the routing algorithm of the sensor network adopts the CRAM algorithm, and the sensors and the wireless gateway adopt a burst wake-up communication mechanism, and only perform wireless network communication when data changes abnormally within a certain time interval or according to the instructions of the environmental monitoring module.

[0094] The embodiments of the present invention can not only timely and accurately monitor the operating environment of escalators and moving walks, but also effectively control the energy consumption of sensor network nodes and extend the service life of the sensor network. Since the sensor network uses a large number of low-cost wireless sensor nodes, compared with wired sensors, it not only greatly saves costs, but also can perform monitoring work more stably in harsh environments.

[0095] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple. For the relevant parts, reference can be made to the description in the method part.

[0096] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An escalator and moving walkway environmental monitoring system based on Internet of Things technology, characterized in that, Including: An environmental monitoring module, which is used to obtain the environmental monitoring data collected by each sensor distributed on the device according to a preset frequency; A data transmission module, which is used to form a sensor network, combine with the General Packet Radio Service (GPRS) technology, remotely transmit the environmental monitoring data collected by each sensor in the target area, and store it in the database of the server; A data analysis module, which is used to perform real-time analysis on the environmental monitoring data collected by each sensor by using a preset model, and determine the warning stage of the environmental monitoring data when the state change trend of the data meets the warning conditions; A warning module, which is used to determine the warning requirements in the warning stage, match the target warning method from the preset warning method library based on the warning requirements to perform warning operations, and send synchronous warning notifications to the management terminal.

2. The escalator and moving walkway environment monitoring system based on Internet of Things technology according to claim 1, characterized in that, It also includes: A visualization module, which is used to display the operation status of the device and the data analysis results in real time, and display warning notifications.

3. The escalator and moving walkway environmental monitoring system based on Internet of Things technology according to claim 1, wherein, The sensors include a temperature sensor, a humidity sensor, a rainfall sensor, and an infrared sensor; The temperature sensor monitors the ambient temperature of the device in real time; The humidity sensor monitors the air humidity of the device in real time; The rainfall sensor monitors the environmental rainfall of the device in real time; The infrared sensor monitors the operation of passengers on the device in real time.

4. The escalator and moving walkway environment monitoring system based on Internet of Things technology according to claim 3, characterized in that, The data analysis module includes: an environmental analysis sub-module, a personnel monitoring sub-module, an anomaly analysis sub-module, and a decision-making sub-module; The environmental analysis sub-module is used to receive the environmental temperature, air humidity, and environmental rainfall data, and generate a temperature change trend chart, a humidity change trend chart, and a rainfall change trend chart respectively; The personnel monitoring sub-module is used to receive the operation of passengers, analyze the peak hours and passenger flow density of personnel, obtain a personnel flow trend chart, and control the device to shut down when no passengers are detected, and control the device to operate normally when passengers are identified; The anomaly analysis sub-module is used to aggregate the environmental temperature, humidity, and rainfall data as inputs, train a machine learning model for prediction, and apply an anomaly detection algorithm to identify abnormal patterns in the data; The decision-making sub-module is used to determine the device operation strategy by combining the results of environmental analysis, personnel monitoring, and anomaly analysis.

5. The escalator and moving walkway environment monitoring system based on the Internet of Things technology according to claim 4, characterized in that, Aggregating the environmental temperature, humidity, and rainfall data as inputs, training a machine learning model for prediction, and applying an anomaly detection algorithm to identify abnormal patterns in the data includes: Obtaining a training sample set, a test sample set, and a feature set; Using the decision tree algorithm to train the training sample set to construct a decision tree. The construction process includes the adaptive threshold of the feature attributes of the training sample set, the selection of the feature attributes of the decision tree node, and the recursive formation of the decision tree based on the feature attribute threshold and the node feature attribute; Using the test sample set to test the decision tree, checking whether the result meets the preset standard. If it meets, use the decision tree to predict the environmental data; If the test result does not meet the preset standard, add the training sample to the training sample set, and re-judge through the above steps that the test result meets the preset standard. Timely adding new training samples to the training sample set is beneficial to adjusting the decision tree that meets the actual classification standard. The new training samples are obtained from the environmental monitoring module; Prune the decision tree to obtain an anomaly recognition model.

6. The escalator and moving walkway environment monitoring system based on the Internet of Things technology according to claim 5, characterized in that, The feature attribute threshold is adaptive, including: for each feature, traverse the feature values and calculate each split point to obtain its split threshold; The selection of the decision tree node feature attributes includes: according to the calculated feature attribute threshold, select the best feature as the splitting feature of the current node; the best feature is the feature with the largest information gain or the smallest Gini index; divide the sample set according to the selected feature attributes and thresholds to form subsets; Recursively form a decision tree, including: for each subset, check whether the stopping condition is met; if the stopping condition is met, mark the current node as a leaf node and assign the corresponding class label; if the stopping condition is not met, repeat the above operations to continue feature selection and division for each subset until the complete decision tree is constructed.

7. The escalator and moving walkway environmental monitoring system based on Internet of Things technology according to claim 6, characterized in that, It also includes: using the grid search method to optimize the decision tree, inputting the monitored environmental data into the optimized decision tree to obtain optimized environmental information for environmental anomaly recognition.

8. The escalator and moving walkway environment monitoring system based on Internet of Things technology according to claim 7, characterized in that, Pruning the decision tree to obtain an anomaly recognition model includes: There are two pruning methods: no pruning and pruning; there are two pruning methods. The error reduction method prunes according to the number of clusters. If the number of clusters is set to 3, randomly select 2 for training, and use the remaining one data to correct the model. The setting of the pruning cluster number is related to the amount of modeling data. If the data volume is greater than 100, set an integer between 2 and 100, otherwise set an integer between 2 and the data volume; Pruning by confidence boundary is to calculate the error estimate value of the node according to the confidence boundary, and then perform pruning, selecting the node with the smallest error rate. The confidence boundary is set to a number between (0, 1); Among them, the method for judging whether pruning is required is: Use the training data set to estimate the error before and after pruning, so as to decide whether to perform pruning: Calculate the upper limit of the confidence level: Among them, N is the number of power data, Pr is the probability density function, E is the number of errors in the electricity consumption category attributes of N power data, q is the true error rate, c is the confidence level, the default value of c is 0.25, and z is the standard deviation corresponding to the confidence level c.

9. The escalator and moving walkway environment monitoring system based on Internet of Things technology according to claim 1, characterized in that, The sensor network is located within the sensing area of the device and communicates with the environmental monitoring module through the communication master station. The sensor network is a ZigBee wireless ad-hoc network, which includes sensors and a wireless gateway. The sensors are used to collect environmental data of escalators and moving walks. The wireless gateway communicates with the sensors wirelessly and communicates with the communication master station through a fieldbus. The wireless gateway transmits the environmental data to the communication master station and transmits the monitoring instructions to the sensors. Among them, the routing algorithm of the sensor network adopts the CRAM algorithm, and the sensors and the wireless gateway adopt a burst wake-up communication mechanism, and only perform wireless network communication when data changes abnormally at certain time intervals or according to the instructions of the environmental monitoring module.

Images