Industrial environment information wireless monitoring system based on Internet of Things and control method thereof

Abstract

The invention discloses an industrial environment information wireless monitoring system based on the Internet of Things and a control method thereof, and relates to the technical field of industrial management, and the system comprises an intelligent edge calculation module, a data visualization module, a self-healing network management module, a precise energy consumption management module, an environment anomaly detection module and a wireless network optimization module. According to the industrial environment information wireless monitoring system based on the Internet of Things and the control method thereof, real-time analysis and decision are realized at a data source by using an intelligent edge computing module, data transmission delay is reduced, response can be quickly made in an emergency, potential faults are predicted through a time sequence abnormity self-learning network model, and the monitoring accuracy of the industrial environment information wireless monitoring system based on the Internet of Things is improved. And network self-healing is actively carried out before a problem occurs, so that the network downtime is effectively reduced. In addition, due to the introduction of a low-power-consumption ad hoc network protocol, the energy consumption optimization management of the sensor nodes is realized, and the service life of equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial management, and in particular to an industrial environment information wireless monitoring system based on the Internet of Things and a control method thereof. Background Art

[0002] With the rapid development of the Industrial Internet of Things (IIoT), industrial environmental information monitoring systems based on wireless sensor networks (WSNs) have been widely used in modern industrial environments. These systems collect various environmental data such as temperature, humidity, vibration, etc. in real time through sensor nodes arranged at industrial sites, and transmit the data to the cloud for analysis and processing. Traditional industrial monitoring systems mainly rely on centralized computing methods, which have problems such as large data transmission delays, heavy computing loads, and slow response speeds. Although edge computing technology has been introduced to solve some of the delay problems, most of them are still at the simple data preprocessing stage and cannot perform more complex intelligent analysis and decision-making on site.

[0003] In the existing wireless sensor network management, the energy consumption of nodes is always a difficult problem. The battery life of sensor nodes is limited, and frequent data transmission will accelerate energy consumption, causing premature node failure and affecting the stability of the entire network. In addition, most existing network management technologies adopt a passive approach, that is, repair operations are performed after the failure occurs, but lack the ability to actively predict failures and quickly self-heal.

[0004] In addition, traditional self-healing network technologies often lack sufficient flexibility in dealing with complex industrial sites. They mostly rely on static routing and fixed algorithms and cannot dynamically adjust according to actual changes in the industrial environment. As for the detection of network anomalies, existing methods rely on the setting of fixed thresholds, which is difficult to effectively respond to changes in the ever-changing industrial environment. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides an industrial environment information wireless monitoring system based on the Internet of Things and a control method thereof to solve the problems raised in the above background technology.

[0006] To achieve the above-mentioned object, the present invention provides the following technical solutions: an industrial environment information wireless monitoring system based on the Internet of Things, comprising an intelligent edge computing module, a data visualization module, a self-healing network management module, a precise energy consumption management module, an environmental anomaly detection module and a wireless network optimization module; The intelligent edge computing module performs real-time computing and processing at the industrial site based on an adaptive learning algorithm. It automatically adjusts the processing algorithm and dynamically adjusts the computing strategy according to the changes in the status of industrial equipment and the real-time characteristics of environmental data at the industrial site to achieve rapid decision support. Data visualization module, through data presentation methods such as 3D models, heat maps, and dynamic charts, managers can intuitively understand the changes in the industrial environment; Self-healing network management module, which is used to automatically detect and repair network connection problems. It uses a timing anomaly self-learning network model to issue warnings when a device or communication link is about to fail, and attempts to reroute data or switch to an alternate communication path to reduce network downtime and ensure the continuity and reliability of data transmission; Precise energy consumption management module, which is used to dynamically adjust energy allocation and usage strategies according to production needs, environmental conditions and real-time data, predict peak energy consumption periods, and schedule non-critical tasks to low energy consumption periods; The environmental anomaly detection module monitors multiple physical parameters in the industrial environment, such as temperature, humidity, air pressure, and air quality, in real time. When a parameter exceeds the set safety range, it quickly identifies the cause of the environmental change and analyzes and predicts the risks that may be caused. The wireless network optimization module uses a low-power ad hoc networking protocol to dynamically adjust the communication path and power consumption strategy according to the distribution of sensor nodes and data transmission requirements to extend battery life and reduce data transmission delays.

[0007] To further optimize the technical solution, in the intelligent edge computing module, a multi-layer dynamic weight adjustment adaptive edge neural network model (MDWA-AENN) is built into the adaptive learning algorithm, and the multi-layer dynamic weight adjustment adaptive edge neural network model includes: Input layer: receives various types of industrial field data, such as temperature, humidity, vibration, and pressure; Feature extraction layer: dynamically adjusts feature weights based on input data to extract data features with high variability; Adaptive adjustment layer: According to the adaptive learning algorithm, the weights and activation functions between neurons are adjusted in real time to adapt to changes in data flow; Decision-making layer: Based on the results of feature extraction and adjustment, the risk value of abnormal changes is calculated, and the response measures to be taken are determined based on the threshold.

[0008] To further optimize the technical solution, in the multi-layer dynamic weight adjustment adaptive edge neural network model: Enter the feature mapping formula in the input layer, assuming the input data is ,in Indicates For each type of sensor data, feature mapping is first performed to standardize and normalize the input data to ensure that each type of data has a similar value range: ; in, and Input data The mean and standard deviation of It is the standardized data; A dynamic weight adjustment mechanism is used in the feature extraction layer to adjust the feature weight according to the changes in the real-time data stream. The weight vector is set as , the weight update formula is: ; in, It's time The weight of the moment; is the learning rate, which is used to control the step size of weight update; It is the gradient of the loss function with respect to the weight, which is used to guide the learning direction; is an adjustment factor used to control the degree of impact according to data changes; Indicates the change in the current input data relative to the previous state; Based on this mechanism, the weights are adjusted when the data stream changes significantly, to ensure that the system responds quickly to data anomalies.

[0009] To further optimize the technical solution, in the multi-layer dynamic weight adjustment adaptive edge neural network model: In the adaptive adjustment layer, the mechanism of adaptive activation function is introduced to improve the adaptability of neural network to different data modes. Different nonlinear functions are dynamically selected according to the changes in input data. The selection formula of the activation function is as follows: ; in, is the change in input data; and It is a preset threshold used to control the switching condition of the activation function; Based on the adaptively selected activation function, the neural network can flexibly adjust its nonlinear mapping ability under different data change scenarios; In the decision-making layer, the system calculates the risk value of abnormal changes based on the results of feature extraction and dynamic adjustment. The risk value calculation formula is as follows: ; in, is the final risk value; It is Standardized input data; is the weight of the corresponding feature; is the importance factor of the feature, obtained from historical data analysis; when Exceeding the set threshold When a fault occurs, the system triggers early warning measures, such as starting equipment protection mechanisms and sending alarm signals for emergency response actions.

[0010] To further optimize the technical solution, the data visualization module, through an interactive graphical interface, allows managers to freely adjust the data display mode when viewing historical data trends, such as zooming in, zooming out, and rotating graphics, to simulate the operation of equipment under different environmental conditions; This module is also based on augmented reality (AR) technology. By wearing AR devices, users can simulate real-time viewing of data points and equipment status collected by various sensors at the industrial site. Managers can intuitively see the operating status, temperature, and humidity information of the equipment at the industrial site, which helps to locate the source of the fault or potential hidden dangers more quickly.

[0011] Further optimizing the technical solution, in the self-healing network management module, the time series anomaly self-learning network model (TASLN) not only focuses on the current changes in the network status, but also predicts the future network status and detects anomalies by learning the historical patterns in the time series, thereby realizing fault prediction and self-healing, and reducing the network downtime of the industrial environment information wireless monitoring system; The timing anomaly self-learning network model includes: Multi-layer temporal convolutional unit: extracts time series features from network data and identifies key state change patterns; Adaptive Long Short-Term Memory Unit (A-LSTM): An adaptive adjustment mechanism is set based on the Long Short-Term Memory Network (LSTM) to adapt to different time series characteristics; Anomaly detection and prediction layer: Based on the dynamic changes of feature patterns, the probability of future failures is calculated through prediction formulas; Self-healing strategy generation unit: dynamically generates and executes self-healing strategies and adjusts data routing or communication paths based on the results of fault prediction.

[0012] To further optimize the technical solution, in the time series anomaly self-learning network model: In the multi-layer time series convolution unit, the features of time series data are extracted. Suppose the input network state sequence is , each Indicates at time The network state, the convolution operation formula is as follows: ; in, is the output feature of the convolutional layer; is the size of the convolution kernel, which indicates the window size in the time dimension; is the weight parameter of the convolution kernel; is the bias term; It is an activation function, such as ReLU, which is used to introduce nonlinear features; Based on multi-layer convolution operations, it effectively captures local features in time series and identifies short-term change patterns of network states. In the adaptive long short-term memory unit, the memory gating mechanism is adaptively adjusted according to the state changes of the time series. The state update formula of A-LSTM is as follows: Input Gate: ; Adaptive forget gate: ; in, Indicates the current state change; It is an adaptive adjustment factor used to amplify or reduce the impact of state changes on the forget gate; Memory unit update: ; Output Gate: ; Hidden state update: ; Based on the adaptive mechanism, LSTM can autonomously adjust the length of memory according to changes in the current network status while processing long-term dependencies, making it more sensitive and efficient when learning network anomalies, ensuring the stability of the Internet of Things environment of the industrial environment information wireless monitoring system.

[0013] To further optimize the technical solution, in the time series anomaly self-learning network model: In the anomaly detection and prediction layer, suppose the future time The failure probability is , the prediction formula is: ; in, For the front The hidden state at a moment; is the time decay factor, used to reduce the impact of distant moments; Indicates the change of the current memory state; is a regulating factor that controls the effect of memory state changes on the probability of failure; By combining historical status features with memory unit changes, the future network status is predicted and the probability of failure is calculated; The self-healing strategy generation unit, when the fault prediction probability Exceeding the set threshold When the system triggers the self-healing strategy generation mechanism, the self-healing strategy generation formula is as follows: ; in, Is the data path A quality score based on network latency, bandwidth, and stability; is the cost of switching paths, such as delay time and resource consumption; is the adjustment factor used to balance path quality and switching cost; Through this formula, the system selects the optimal communication path before a failure occurs, ensuring the continuity and reliability of data transmission.

[0014] To further optimize the technical solution, in the wireless network optimization module, the low-power self-organizing network protocol includes the following contents: Dynamic node wake-up mechanism: A dynamic node wake-up mechanism based on data transmission needs is adopted. The node will switch from low-power mode to active state only when it needs to transmit or receive data; Multi-hop routing optimization: Dynamically adjust the multi-hop transmission path of data by analyzing the distribution of sensor nodes in the network; Adaptive power consumption control: intelligently adjusts the transmit power of each node based on the remaining power of the node and the current network load; Cluster head node self-selection mechanism: The sensor nodes in the network periodically elect cluster head nodes to centrally process and forward data.

[0015] A control method of an industrial environment information wireless monitoring system based on the Internet of Things is operated based on the above-mentioned industrial environment information wireless monitoring system based on the Internet of Things, and includes the following specific steps: S1, data collection and transmission start; Start sensor nodes to collect various data in industrial environments, such as temperature, humidity, vibration, and pressure, and transmit them to edge devices through wireless sensor networks; S2, intelligent edge computing; Process and analyze data in real time, filter out noisy data, focus on abnormal data points, and conduct preliminary risk assessment; S3, Fault prediction analysis; Use the time series anomaly self-learning network model to analyze historical data of network status, predict possible network failures or data transmission problems, and calculate the failure probability; S4, dynamic optimization of communication paths; Based on the low-power self-organizing network protocol, dynamically adjust the data transmission path in the wireless sensor network, optimize communication efficiency, reduce node energy consumption, and extend the service life of the equipment; S5, real-time response and control execution; Based on the results of edge computing and fault prediction, corresponding control measures are implemented for possible risks or faults, such as rerouting data, activating backup devices, or adjusting sensor node status; S6, self-healing mechanism triggered; When the predicted probability of failure exceeds the set threshold, the self-healing mechanism is automatically triggered to reconfigure the network routing or switch to an alternative communication path to ensure the continuity and stability of data transmission; S7, remote monitoring and feedback adjustment; The system uploads the processed data and device status to the cloud for remote monitoring and analysis by managers, and dynamically adjusts the working parameters of edge devices and sensor nodes based on cloud feedback; S8, energy consumption management and optimization; Regularly use the adaptive power consumption control mechanism in the low-power ad hoc network protocol to adjust the transmission power of each sensor node and optimize the energy consumption strategy according to the node power and communication distance; S9, fault record and analysis report; After each fault prediction or self-healing operation, the system automatically generates event records and analysis reports for subsequent data analysis and improvement reference.

[0016] Compared with the prior art, the present invention provides an industrial environment information wireless monitoring system based on the Internet of Things and a control method thereof, which has the following beneficial effects: This industrial environment information wireless monitoring system based on the Internet of Things and its control method, by utilizing an intelligent edge computing module, realizes real-time analysis and decision-making at the data source, reduces data transmission delays, and can respond quickly in emergency situations. It predicts potential faults through a timing anomaly self-learning network model, and actively performs network self-healing before problems occur, effectively reducing network downtime.

[0017] In addition, the introduction of low-power self-organizing network protocols has achieved optimal energy consumption management of sensor nodes, extended the service life of equipment, and enabled the system to significantly improve monitoring reliability, response speed, and data transmission stability when facing complex and changing industrial environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a structural schematic diagram of an industrial environment information wireless monitoring system based on the Internet of Things proposed by the present invention; Figure 2 A schematic diagram of the process of the MDWA-AENN model in an industrial environment information wireless monitoring system based on the Internet of Things proposed by the present invention; Figure 3 A flow chart of the TASLN model in an industrial environment information wireless monitoring system based on the Internet of Things proposed by the present invention; Figure 4 The present invention is a flowchart of a control method for an industrial environment information wireless monitoring system based on the Internet of Things. DETAILED DESCRIPTION

[0019] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Embodiment 1: See also Figure 1 , an industrial environment information wireless monitoring system based on the Internet of Things, includes an intelligent edge computing module, a data visualization module, a self-healing network management module, a precise energy consumption management module, an environmental anomaly detection module and a wireless network optimization module.

[0021] The intelligent edge computing module performs real-time computing and processing at the industrial site based on an adaptive learning algorithm. It automatically adjusts the processing algorithm and dynamically adjusts the computing strategy according to the changes in the status of industrial equipment and the real-time characteristics of environmental data at the industrial site to achieve rapid decision support.

[0022] In this embodiment, in the intelligent edge computing module, a multi-layer dynamic weight adjustment adaptive edge neural network model (MDWA-AENN) is built into the adaptive learning algorithm. The multi-layer dynamic weight adjustment adaptive edge neural network model includes: Input layer: receives various types of industrial field data, such as temperature, humidity, vibration, and pressure; Feature extraction layer: dynamically adjusts feature weights based on input data to extract data features with high variability; Adaptive adjustment layer: According to the adaptive learning algorithm, the weights and activation functions between neurons are adjusted in real time to adapt to changes in data flow; Decision-making layer: Based on the results of feature extraction and adjustment, the risk value of abnormal changes is calculated, and the response measures to be taken are determined based on the threshold.

[0023] Furthermore, in the multi-layer dynamic weight adjustment adaptive edge neural network model: Enter the feature mapping formula in the input layer, assuming the input data is ,in Indicates For each type of sensor data, feature mapping is first performed to standardize and normalize the input data to ensure that each type of data has a similar value range: ; in, and Input data The mean and standard deviation of It is the standardized data; A dynamic weight adjustment mechanism is used in the feature extraction layer to adjust the feature weight according to the changes in the real-time data stream. The weight vector is set as , the weight update formula is: ; in, It's time The weight of the moment; is the learning rate, which is used to control the step size of weight update; It is the gradient of the loss function with respect to the weight, which is used to guide the learning direction; is an adjustment factor used to control the degree of impact according to data changes; Indicates the change in the current input data relative to the previous state; Based on this mechanism, the weights are adjusted when the data stream changes significantly, to ensure that the system responds quickly to data anomalies.

[0024] In the adaptive adjustment layer, the mechanism of adaptive activation function is introduced to improve the adaptability of neural network to different data modes. Different nonlinear functions are dynamically selected according to the changes in input data. The selection formula of the activation function is as follows: ; in, is the change in input data; and It is a preset threshold used to control the switching condition of the activation function; Based on the adaptively selected activation function, the neural network can flexibly adjust its nonlinear mapping ability under different data change scenarios; In the decision-making layer, the system calculates the risk value of abnormal changes based on the results of feature extraction and dynamic adjustment. The risk value calculation formula is as follows: ; in, is the final risk value; It is Standardized input data; is the weight of the corresponding feature; is the importance factor of the feature, obtained from historical data analysis; when Exceeding the set threshold When a fault occurs, the system triggers early warning measures, such as starting equipment protection mechanisms and sending alarm signals for emergency response actions.

[0025] like Figure 2 As shown, the model, when used, includes: Data preprocessing: The system first receives a variety of sensor data from the industrial site and standardizes it by inputting feature mapping formulas to ensure that the numerical range of each type of data is consistent.

[0026] Dynamic weight update: According to the changes in real-time data streams, the model uses a dynamic weight adjustment mechanism to adjust feature weights through gradient information and data changes to increase the focus on important features.

[0027] Activation function selection: In the feature extraction stage, the model adaptively selects the appropriate activation function according to the changes in the current data to improve its adaptability to complex pattern changes.

[0028] Risk assessment and response: The decision-making layer comprehensively considers the risk impact of all features and calculates the risk value through the risk assessment formula. Based on the size of the risk value, it determines whether it is necessary to trigger an emergency response or adjust the operation strategy.

[0029] The system can quickly adapt to data changes under different environmental conditions. At the same time, through real-time analysis and prediction of industrial environmental data, ultra-low latency edge computing capabilities are achieved, significantly improving the system's response speed and accuracy in emergency situations.

[0030] The data visualization module enables managers to intuitively understand changes in the industrial environment through data presentation methods such as 3D models, heat maps, and dynamic charts.

[0031] In this embodiment, the data visualization module, through an interactive graphical interface, allows managers to freely adjust the data display mode when viewing historical data trends, such as zooming in, zooming out, and rotating graphics, to simulate the operation of equipment under different environmental conditions; This module is also based on augmented reality (AR) technology. By wearing AR devices, users can simulate real-time viewing of data points and equipment status collected by various sensors at the industrial site. Managers can intuitively see the operating status, temperature, and humidity information of the equipment at the industrial site, which helps to locate the source of the fault or potential hidden dangers more quickly.

[0032] The self-healing network management module is used to automatically detect and repair network connection problems. It uses a fault prediction algorithm to issue a warning when a device or communication link is about to fail, and attempts to reroute data or switch to an alternate communication path to reduce network downtime and ensure the continuity and reliability of data transmission.

[0033] In this embodiment, in the self-healing network management module, the fault prediction algorithm has a built-in time series anomaly self-learning network model (TASLN), which not only focuses on the current changes in the network state, but also predicts and detects anomalies in the future network state by learning historical patterns in the time series, thereby realizing fault prediction and self-healing, and reducing the network downtime of the industrial environment information wireless monitoring system; The timing anomaly self-learning network model includes: Multi-layer temporal convolutional unit: extracts time series features from network data and identifies key state change patterns; Adaptive Long Short-Term Memory (A-LSTM): An adaptive adjustment mechanism is set based on LSTM to adapt to different timing characteristics; Anomaly detection and prediction layer: Based on the dynamic changes of feature patterns, the probability of future failures is calculated through prediction formulas; Self-healing strategy generation unit: dynamically generates and executes self-healing strategies and adjusts data routing or communication paths based on the results of fault prediction.

[0034] Furthermore, in the time series anomaly self-learning network model: In the multi-layer time series convolution unit, the features of time series data are extracted. Suppose the input network state sequence is , each Indicates at time The network state, the convolution operation formula is as follows: ; in, is the output feature of the convolutional layer; is the size of the convolution kernel, which indicates the window size in the time dimension; is the weight parameter of the convolution kernel; is the bias term; It is an activation function, such as ReLU, which is used to introduce nonlinear features; Based on multi-layer convolution operations, it effectively captures local features in time series and identifies short-term change patterns of network states. In the adaptive long short-term memory unit, the memory gating mechanism is adaptively adjusted according to the state changes of the time series. The state update formula of A-LSTM is as follows: Input Gate: ; Adaptive forget gate: ; in, Indicates the current state change; It is an adaptive adjustment factor used to amplify or reduce the impact of state changes on the forget gate; Memory unit update: ; Output Gate: ; Hidden state update: ; Based on the adaptive mechanism, LSTM can autonomously adjust the length of memory according to changes in the current network status while processing long-term dependencies, making it more sensitive and efficient when learning network anomalies, ensuring the stability of the Internet of Things environment of the industrial environment information wireless monitoring system.

[0035] In the anomaly detection and prediction layer, suppose the future time The failure probability is , the prediction formula is: ; in, For the front The hidden state at a moment; is the time decay factor, used to reduce the impact of distant moments; Indicates the change of the current memory state; is a regulating factor that controls the effect of memory state changes on the probability of failure; By combining historical status features with memory unit changes, the future network status is predicted and the probability of failure is calculated; The self-healing strategy generation unit, when the fault prediction probability Exceeding the set threshold When the system triggers the self-healing strategy generation mechanism, the self-healing strategy generation formula is as follows: ; in, Is the data path A quality score based on network latency, bandwidth, and stability; is the cost of switching paths, such as delay time and resource consumption; is the adjustment factor used to balance path quality and switching cost; Through this formula, the system selects the optimal communication path before a failure occurs, ensuring the continuity and reliability of data transmission.

[0036] like Figure 3 As shown, the model, when used, includes: Time series feature extraction: The model first extracts features from historical network state data through multiple layers of time series convolutional units to identify local time series patterns.

[0037] Adaptive state update: Use adaptive long short-term memory unit (A-LSTM) to autonomously adjust the state of the memory unit according to the short-term change of the network state to accurately remember the long-term and short-term dependency information.

[0038] Fault probability prediction: Through the prediction formula of the anomaly detection and fault prediction layer, the probability of fault occurrence in the future is calculated, and the health status of the network connection is evaluated in real time.

[0039] Self-healing strategy generation: If the failure probability exceeds the set threshold, the system will dynamically generate the optimal self-healing strategy based on the path quality and switching cost, and automatically adjust the communication path to avoid data interruption.

[0040] The system can not only identify complex time series features, but also flexibly adjust model parameters to accurately predict faults according to real-time status changes. The design of the self-healing strategy generation mechanism enables the system to respond quickly and automatically repair faults, greatly improving the self-healing ability and reliability of network management.

[0041] The precise energy consumption management module is used to dynamically adjust energy distribution and usage strategies according to production needs, environmental conditions and real-time data, predict peak energy consumption periods, and schedule non-critical tasks to low energy consumption periods.

[0042] In this embodiment, for the multi-energy modes of large factories (such as electricity, natural gas, solar energy, etc.), the module can analyze the efficiency of various energy sources in real time and provide optimal energy consumption strategy recommendations. This can not only significantly reduce energy costs, but also reduce carbon emissions, helping to achieve the goal of green factories.

[0043] The environmental anomaly detection module monitors multiple physical parameters in the industrial environment, such as temperature, humidity, air pressure, and air quality, in real time. When a parameter exceeds the set safety range, it quickly identifies the cause of the environmental change and analyzes and predicts the risks that may be caused.

[0044] In this embodiment, the change patterns of multiple environmental parameters are comprehensively analyzed instead of relying solely on the fluctuation of a certain parameter. For example, a temperature increase may be accompanied by a decrease in humidity and a change in air pressure. The linkage information of these parameters can help to more accurately determine the source of the problem.

[0045] The wireless network optimization module uses a low-power ad hoc networking protocol to dynamically adjust the communication path and power consumption strategy according to the distribution of sensor nodes and data transmission requirements to extend battery life and reduce data transmission delays.

[0046] In this embodiment, in the wireless network optimization module, the low-power self-organizing network protocol includes the following contents: Dynamic node wake-up mechanism: A dynamic node wake-up mechanism based on data transmission needs is adopted. The node will switch from low-power mode to active state only when it needs to transmit or receive data. This effectively reduces energy consumption and avoids nodes being in a power-consuming state during unnecessary periods of time.

[0047] Multi-hop routing optimization: By analyzing the distribution of sensor nodes in the network, the multi-hop transmission path of data is dynamically adjusted; the path with the lowest power consumption and better communication quality is given priority to reduce data transmission delays, and the energy consumption of each node is balanced to prevent individual nodes from running out of power prematurely.

[0048] Adaptive power consumption control: Intelligently adjust the transmit power of each node according to the remaining power of the node and the current network load; reduce the transmit power to save energy for communications at a closer distance; and appropriately increase the transmit power to ensure the reliability of data transmission for critical data or nodes at a longer distance.

[0049] Cluster head node self-selection mechanism: The sensor nodes in the network periodically elect cluster head nodes to centrally process and forward data. The election process takes into account factors such as the remaining power of the node and the communication quality to ensure the efficiency and durability of the cluster head node, thereby reducing the overall network energy consumption.

[0050] Embodiment 2: See also Figure 4 A control method for an industrial environment information wireless monitoring system based on the Internet of Things is operated based on the industrial environment information wireless monitoring system based on the Internet of Things described in Example 1, and includes the following specific steps: S1. Data collection and transmission start Start sensor nodes to collect various data in industrial environments, such as temperature, humidity, vibration, and pressure, and transmit them to edge devices through wireless sensor networks.

[0051] S2, Intelligent Edge Computing Process and analyze data in real time, filter out noisy data, focus on abnormal data points, and conduct preliminary risk assessment.

[0052] S3. Fault prediction analysis Use the time series anomaly self-learning network model to analyze historical data of network status, predict possible network failures or data transmission problems, and calculate the failure probability.

[0053] S4. Dynamic optimization of communication paths Based on the low-power self-organizing network protocol, the data transmission path in the wireless sensor network is dynamically adjusted to optimize the communication efficiency, reduce the node energy consumption, and extend the service life of the equipment.

[0054] S5. Real-time response and control execution Based on the results of edge computing and fault prediction, corresponding control measures are performed for possible risks or failures, such as rerouting data, activating backup equipment, or adjusting sensor node status.

[0055] S6. Self-healing mechanism triggered When the predicted probability of failure exceeds the set threshold, the self-healing mechanism is automatically triggered to reconfigure the network routing or switch to an alternative communication path to ensure the continuity and stability of data transmission.

[0056] S7, Remote monitoring and feedback adjustment The system uploads the processed data and device status to the cloud for management personnel to remotely monitor and analyze, and dynamically adjust the working parameters of edge devices and sensor nodes based on cloud feedback.

[0057] S8. Energy consumption management and optimization The adaptive power consumption control mechanism in the low-power ad hoc network protocol is regularly used to adjust the transmission power of each sensor node and optimize the energy consumption strategy according to the node power and communication distance.

[0058] S9. Fault Record and Analysis Report After each fault prediction or self-healing operation, the system automatically generates event records and analysis reports for subsequent data analysis and improvement reference.

[0059] The beneficial effects of the present invention are: This industrial environment information wireless monitoring system based on the Internet of Things and its control method, by utilizing an intelligent edge computing module, realizes real-time analysis and decision-making at the data source, reduces data transmission delays, and can respond quickly in emergency situations. It predicts potential faults through a timing anomaly self-learning network model, and actively performs network self-healing before problems occur, effectively reducing network downtime.

[0060] In addition, the introduction of low-power self-organizing network protocols has achieved optimal energy consumption management of sensor nodes, extended the service life of equipment, and enabled the system to significantly improve monitoring reliability, response speed, and data transmission stability when facing complex and changing industrial environments.

[0061] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0062] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An industrial environment information wireless monitoring system based on the Internet of Things, characterized in that: It includes intelligent edge computing module, data visualization module, self-healing network management module, precise energy consumption management module, environmental anomaly detection module and wireless network optimization module; The intelligent edge computing module performs real-time computing and processing at the industrial site based on an adaptive learning algorithm. It automatically adjusts the processing algorithm and dynamically adjusts the computing strategy according to the changes in the status of industrial equipment and the real-time characteristics of environmental data at the industrial site to achieve rapid decision support. Data visualization module, through data presentation methods including 3D models, heat maps, and dynamic charts, managers can intuitively understand the changes in the industrial environment; Self-healing network management module, which is used to automatically detect and repair network connection problems. It uses a timing anomaly self-learning network model to issue warnings when a device or communication link is about to fail, and attempts to reroute data or switch to an alternate communication path to reduce network downtime and ensure the continuity and reliability of data transmission; Precise energy consumption management module, which is used to dynamically adjust energy allocation and usage strategies according to production needs, environmental conditions and real-time data, predict peak energy consumption periods, and schedule non-critical tasks to low energy consumption periods; The environmental anomaly detection module monitors multiple physical parameters in the industrial environment, including temperature, humidity, air pressure, and air quality, in real time. When a parameter exceeds the set safety range, it quickly identifies the cause of the environmental change and analyzes and predicts the risks that may be caused. The wireless network optimization module uses a low-power ad hoc networking protocol to dynamically adjust the communication path and power consumption strategy according to the distribution of sensor nodes and data transmission requirements to extend battery life and reduce data transmission delays.

2. According to the Internet of Things-based industrial environment information wireless monitoring system of claim 1, it is characterized in that: In the intelligent edge computing module, a multi-layer dynamic weight adjustment adaptive edge neural network model is built into the adaptive learning algorithm, and the multi-layer dynamic weight adjustment adaptive edge neural network model includes: Input layer: receives various types of industrial field data, including temperature, humidity, vibration, and pressure; Feature extraction layer: dynamically adjusts feature weights based on input data to extract data features with high variability; Adaptive adjustment layer: According to the adaptive learning algorithm, the weights and activation functions between neurons are adjusted in real time to adapt to changes in data flow; Decision-making layer: Based on the results of feature extraction and adjustment, the risk value of abnormal changes is calculated, and the response measures to be taken are determined based on the threshold.

3. According to the Internet of Things-based industrial environment information wireless monitoring system of claim 2, it is characterized in that: In the multi-layer dynamic weight adjustment adaptive edge neural network model: Enter the feature mapping formula in the input layer, assuming the input data is ,in Indicates For each type of sensor data, feature mapping is first performed to standardize and normalize the input data to ensure that each type of data has a similar value range: ； in, and Input data The mean and standard deviation of It is the standardized data; A dynamic weight adjustment mechanism is used in the feature extraction layer to adjust the feature weight according to the changes in the real-time data stream. The weight vector is set as , the weight update formula is: ； in, It's time The weight of the moment; is the learning rate, which is used to control the step size of weight update; It is the gradient of the loss function with respect to the weight, which is used to guide the learning direction; is an adjustment factor used to control the degree of impact according to data changes; Indicates the change in the current input data relative to the previous state; Based on this mechanism, the weights are adjusted when the data flow changes, ensuring that the system responds quickly to data anomalies.

4. According to the Internet of Things-based industrial environment information wireless monitoring system of claim 2, it is characterized in that: In the multi-layer dynamic weight adjustment adaptive edge neural network model: In the adaptive adjustment layer, the mechanism of adaptive activation function is introduced to improve the adaptability of neural network to different data modes. Different nonlinear functions are dynamically selected according to the changes in input data. The selection formula of the activation function is as follows: ； in, is the change in input data; and It is a preset threshold used to control the switching condition of the activation function; Based on the adaptively selected activation function, the neural network can flexibly adjust its nonlinear mapping ability under different data change scenarios; In the decision-making layer, the system calculates the risk value of abnormal changes based on the results of feature extraction and dynamic adjustment. The risk value calculation formula is as follows: ； in, is the final risk value; It is Standardized input data; is the weight of the corresponding feature; is the importance factor of the feature, obtained from historical data analysis; when Exceeding the set threshold When a fault occurs, the system triggers early warning measures, including emergency response actions such as starting equipment protection mechanisms and sending alarm signals.

5. According to the Internet of Things-based industrial environment information wireless monitoring system of claim 1, it is characterized in that: The data visualization module, through an interactive graphical interface, allows managers to freely adjust the data display method when viewing historical data trends, including zooming in, zooming out, and rotating graphics to simulate the operation of equipment under different environmental conditions; This module is also based on augmented reality technology. By wearing augmented reality devices, users can simulate real-time viewing of data points and equipment status collected by various sensors at the industrial site. Managers can intuitively see the operating status, temperature, and humidity information of the equipment at the industrial site, which helps to locate the source of the fault or potential hidden dangers more quickly.

6. The industrial environment information wireless monitoring system based on the Internet of Things according to claim 1 is characterized in that: In the self-healing network management module, the time series anomaly self-learning network model not only focuses on the current changes in the network status, but also predicts the future network status and detects anomalies by learning the historical patterns in the time series, thereby realizing fault prediction and self-healing, and reducing the network downtime of the industrial environment information wireless monitoring system; The timing anomaly self-learning network model includes: Multi-layer temporal convolutional unit: extracts time series features from network data and identifies key state change patterns; Adaptive long short-term memory unit: An adaptive adjustment mechanism is set based on the long short-term memory network to adapt to different temporal characteristics; Anomaly detection and prediction layer: Based on the dynamic changes of feature patterns, the probability of future failures is calculated through prediction formulas; Self-healing strategy generation unit: dynamically generates and executes self-healing strategies and adjusts data routing or communication paths based on the results of fault prediction.

7. The industrial environment information wireless monitoring system based on the Internet of Things according to claim 6 is characterized in that: In the timing anomaly self-learning network model: In the multi-layer time series convolution unit, the features of time series data are extracted. Suppose the input network state sequence is , each Indicates at time The network state, the convolution operation formula is as follows: ； in, is the output feature of the convolutional layer; is the size of the convolution kernel, which indicates the window size in the time dimension; is the weight parameter of the convolution kernel; is the bias term; It is an activation function, including ReLU, which is used to introduce nonlinear features; Based on multi-layer convolution operations, it effectively captures local features in time series and identifies short-term change patterns of network states. In the adaptive long short-term memory unit, the memory gating mechanism is adaptively adjusted according to the state change of the time series. The state update formula of the adaptive long short-term memory unit is as follows: Input Gate: ； Adaptive forget gate: ； in, Indicates the current state change; It is an adaptive adjustment factor used to amplify or reduce the impact of state changes on the forget gate; Memory unit update: ； Output Gate: ； Hidden state update: ； Based on the adaptive mechanism, the long short-term memory network can process long-term dependencies while autonomously adjusting the length of memory according to changes in the current network status. It is more sensitive and efficient when learning network anomalies, ensuring the stability of the Internet of Things environment of the industrial environment information wireless monitoring system.

8. The industrial environment information wireless monitoring system based on the Internet of Things according to claim 6 is characterized in that: In the timing anomaly self-learning network model: In the anomaly detection and prediction layer, suppose the future time The failure probability is , the prediction formula is: ； in, For the front The hidden state at a moment; is the time decay factor, used to reduce the influence of the previous moments; Indicates the change of the current memory state; is a regulating factor that controls the effect of memory state changes on the probability of failure; By combining historical status features with memory unit changes, the future network status is predicted and the probability of failure is calculated; The self-healing strategy generation unit, when the fault prediction probability Exceeding the set threshold When the system triggers the self-healing strategy generation mechanism, the self-healing strategy generation formula is as follows: ； in, Is the data path A quality score based on network latency, bandwidth, and stability; is the cost of the switching path, including delay time and resource consumption; is the adjustment factor used to balance path quality and switching cost; Through this formula, the system selects the optimal communication path before a failure occurs, ensuring the continuity and reliability of data transmission.

9. The industrial environment information wireless monitoring system based on the Internet of Things according to claim 1 is characterized in that: In the wireless network optimization module, the low-power self-organizing network protocol includes the following contents: Dynamic node wake-up mechanism: A dynamic node wake-up mechanism based on data transmission needs is adopted. The node will switch from low-power mode to active state only when it needs to transmit or receive data; Multi-hop routing optimization: Dynamically adjust the multi-hop transmission path of data by analyzing the distribution of sensor nodes in the network; Adaptive power consumption control: intelligently adjusts the transmit power of each node based on the remaining power of the node and the current network load; Cluster head node self-selection mechanism: The sensor nodes in the network periodically elect cluster head nodes to centrally process and forward data.

10. A control method for an industrial environment information wireless monitoring system based on the Internet of Things, operating based on an industrial environment information wireless monitoring system based on the Internet of Things according to any one of claims 1 to 9, characterized in that: The specific steps include: S1, data collection and transmission start; Start sensor nodes to collect various data in industrial environments, including temperature, humidity, vibration, and pressure, and transmit them to edge devices through wireless sensor networks; S2, intelligent edge computing; Process and analyze data in real time, filter out noisy data, focus on abnormal data points, and conduct preliminary risk assessment; S3, Fault prediction analysis; Use the time series anomaly self-learning network model to analyze historical data of network status, predict possible network failures or data transmission problems, and calculate the failure probability; S4, dynamic optimization of communication paths; Based on the low-power self-organizing network protocol, dynamically adjust the data transmission path in the wireless sensor network, optimize communication efficiency, reduce node energy consumption, and extend the service life of the equipment; S5, real-time response and control execution; Based on the results of edge computing and fault prediction, corresponding control measures are implemented for possible risks or faults, including restarting routing data, activating backup devices, or adjusting sensor node status; S6, self-healing mechanism triggered; When the predicted probability of failure exceeds the set threshold, the self-healing mechanism is automatically triggered to reconfigure the network routing or switch to an alternative communication path to ensure the continuity and stability of data transmission; S7, remote monitoring and feedback adjustment; The system uploads the processed data and device status to the cloud for remote monitoring and analysis by managers, and dynamically adjusts the working parameters of edge devices and sensor nodes based on cloud feedback; S8, energy consumption management and optimization; Regularly use the adaptive power consumption control mechanism in the low-power ad hoc network protocol to adjust the transmission power of each sensor node and optimize the energy consumption strategy according to the node power and communication distance; S9, fault record and analysis report; After each fault prediction or self-healing operation, the system automatically generates event records and analysis reports for subsequent data analysis and improvement reference.

Images