Environmental protection equipment management and data analysis system based on Internet of Things

Through the combination of IoT sensor groups and adaptive algorithms, the intelligent, dynamic adaptation and real-time decision-making of environmentally friendly equipment are achieved, the problem of lag in fault prediction in traditional systems is solved, and the efficiency and reliability of equipment management are improved.

CN120277603AInactive Publication Date: 2025-07-08深圳市利濠城市服务有限公司
View PDF 0 Cites 3 Cited by

Patent Information

Application Number
CN202510212340.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-02-25
Publication Date
2025-07-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional environmental protection equipment management systems lack intelligent, dynamic adaptability and data-driven real-time decision-making support, resulting in lagging equipment failure prediction and high maintenance costs, making it difficult to achieve efficient management.

Method used

The Internet of Things-based sensor group, central data processing module and cloud database system are adopted, combined with adaptive algorithms and multi-dimensional data analysis, and the sampling frequency and transmission path are dynamically adjusted to perform real-time monitoring and fault prediction.

Benefits of technology

Improve the monitoring accuracy of environmentally friendly equipment, optimize the acquisition strategy, enhance the fault prediction capability, and significantly reduce downtime and maintenance costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120277603A_ABST
    Figure CN120277603A_ABST
Patent Text Reader

Abstract

The invention discloses an environmental protection equipment management and data analysis system based on the Internet of Things. The environmental protection equipment management and data analysis system comprises an Internet of Things sensor group, a central data processing module and a cloud database, the operation process of the system specifically comprises the following steps that the Internet of Things sensor groups are installed on the multiple pieces of environmental protection equipment to comprehensively collect data; the Internet of Things sensor group transmits acquired data to the central data processing module; the central data processing module performs data preprocessing on the original data; performing multi-dimensional analysis and calculation on the preprocessed data to output a fault prediction report; and uploading all data to a cloud database for storage and management. The method has the following advantages and effects: the defects of the existing environmental protection equipment in the aspects of intelligence, dynamic adaptation, data processing and real-time decision making are improved, the equipment monitoring precision can be improved, the acquisition strategy can be optimized, the fault prediction capability can be enhanced, and the downtime and the maintenance cost can be remarkably reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of environmental protection equipment management, and particularly to an environmental protection equipment management and data analysis system based on the Internet of Things. Background Art

[0002] With the increasing awareness of environmental protection, environmental protection facilities in industries and cities play a crucial role in improving air quality, reducing pollutant emissions, etc. Traditional environmental protection equipment management and operation monitoring mainly rely on manual monitoring and regular maintenance, suffering from problems such as untimely collection of monitoring data, single management means, and lagging equipment fault prediction, making it difficult to achieve comprehensive monitoring and efficient management of equipment status. Therefore, an environmental protection equipment management and data analysis system based on the Internet of Things has emerged, using advanced information technology, Internet of Things technology, artificial intelligence technology, and big data analysis means to provide support for the efficient management of environmental protection equipment; however, the main problems faced in actual performance are the lack of intelligence, dynamic adaptation ability, and data-driven real-time decision-making support. Traditional systems rely on fixed sampling frequencies, static monitoring methods, and manual intervention, resulting in the inability to adjust the acquisition strategy and response mechanism in a timely manner when equipment loads change, environmental fluctuations occur, or equipment failures happen. At the same time, there is a lack of sufficient intelligent analysis capabilities, making it difficult to perform accurate fault prediction, early warning, and optimization control based on data and historical trends, resulting in equipment maintenance relying on post-event diagnosis, increasing downtime and maintenance costs. Therefore, the deficiencies of the existing technology in terms of intelligence, dynamic adaptation, data processing, and real-time decision-making limit the efficient operation and sustainable development of environmental protection equipment management systems. Summary of the Invention

[0003] The purpose of the present invention is to provide an environmental protection equipment management and data analysis system based on the Internet of Things to solve the problems raised in the background art.

[0004] The above technical purpose of the present invention is achieved through the following technical solutions:

[0005] An environmental protection equipment management and data analysis system based on the Internet of Things includes an Internet of Things sensor group installed on each environmental protection equipment, a central data processing module that realizes data intercommunication with the Internet of Things sensor groups on each environmental protection equipment through Internet of Things communication, and a cloud database that can receive data from the central data processing module; the specific operation process of the system includes the following steps:

[0006] S100. Comprehensively collect data through the Internet of Things sensor groups on each environmental protection equipment, and the Internet of Things sensor group includes multiple different types of sensors;

[0007] S200. Each Internet of Things sensor group installed on the environmental protection equipment transmits the collected data to the central data processing module through the Internet of Things communication network;

[0008] S300. After the central data processing module receives the data transmitted by each Internet of Things sensor, it preprocesses the data.

[0009] S400. Perform multi-dimensional analysis and calculation on the preprocessed data to output a fault prediction report to help management personnel timely understand the operating status and potential risks of environmental protection equipment.

[0010] S500. Upload all the data from step S100 to step S400 to the cloud database for storage and management.

[0011] By adopting the above technical solution, by installing multiple different types of sensors on each environmental protection equipment, the system can comprehensively monitor the operating conditions of the equipment, including working parameters and external environmental factors; the environmental protection equipment transmits the collected data to the central data processing module through the Internet of Things communication network to form centralized management of the data. Through centralized data processing, the data of the equipment can be obtained and processed in a timely manner to ensure the health and normal operation of each environmental protection equipment; by real-time monitoring the working status of the environmental protection equipment and combining with external environmental changes, the system can timely identify potential equipment failures, thereby helping management personnel timely understand the operating status and potential risks of the equipment.

[0012] A further setting is that in step S100, each sensor collects data at regular intervals according to a preset sampling period. The sampling period is dynamically adjusted based on monitoring requirements. During the dynamic adjustment process, the sampling frequency is automatically adjusted according to the load status, battery energy efficiency, and external environmental variable factors of the environmental protection equipment through an adaptive algorithm to balance the real-time performance and energy efficiency of data collection.

[0013] By adopting the above technical solution, the adaptive algorithm adjusts the sampling frequency of the sensor by real-time calculating and analyzing the working load of the environmental protection equipment, the energy efficiency of the equipment battery, and the changes in external environmental variables, so as to achieve the purposes of reducing energy consumption, better coping with sudden environmental changes, avoiding data distortion caused by environmental interference, and being able to give priority to ensuring the operation of key sensors when the battery power is insufficient to avoid data interruption caused by power exhaustion, and ultimately improving the system robustness.

[0014] A further setting is that the specific steps for the adaptive algorithm to automatically adjust the sampling frequency are as follows:

[0015] S101. Make a first adjustment to the sampling frequency based on the load status of the environmental protection equipment:

[0016] f L (t)=f0×(1 + α·L(t))

[0017] where, f L(t) is the impact of the device load on the sampling frequency, f0 is the default sampling frequency, α is the load adjustment factor, L(t) is the current load status of the environmental protection device, and t is the timestamp;

[0018] S102. Under the influence of the load status of the environmental protection device, the sampling frequency is adjusted secondarily based on the battery energy efficiency of the environmental protection device:

[0019] f η (t) = f L (t) × (1 - β·η(t))

[0020] Among them, f η (t) is the impact of the battery energy efficiency of the environmental protection device on the sampling frequency, β is the energy efficiency adjustment factor, and η(t) is the battery energy consumption of the current environmental protection device;

[0021] S103. Under the influence of the battery energy efficiency of the environmental protection device, the sampling frequency is adjusted thirdly based on the external environmental variables:

[0022] f E (t) = f η (t) × (1 + γ·E(t))

[0023] Among them, f E (t) is the impact of the external environmental variables on the sampling frequency, γ is the environmental adjustment factor, and E(t) is the current external environmental variable;

[0024] S104. After the sampling frequency is adjusted three times, based on the weighted average of the load status, battery energy efficiency, and external environmental variable factors of the environmental protection device, the dynamic sampling frequency f dynamic (t) is:

[0025] f dynamic (t) = w1·f L (t) + w2·f η (t) + w3·f E (t)

[0026] Among them, w1, w2, and w3 are the weight coefficients of the load status, battery energy efficiency, and external environmental variables of the environmental protection device respectively.

[0027] By adopting the above technical solution, the load status directly affects the safety of environmental protection equipment. The sampling frequency is adjusted once to quickly respond to load changes. In the second adjustment of the sampling frequency, the battery energy consumption factor β is introduced to prevent power-off of the equipment caused by over-sampling. In the third adjustment of the sampling frequency, it is dynamically corrected through the environmental adjustment factor γ. Through the adaptive adjustment of the sampling frequency, especially in the case of low load or low battery power, the system can reduce the data acquisition volume, thereby saving energy consumption, and maintain a sufficient data update frequency while ensuring the low-energy consumption operation of the equipment, improving the system response speed and the working efficiency of the equipment.

[0028] A further setting is that in the step S200, it is necessary to select the best network path in the Internet of Things communication network to improve data transmission efficiency and reduce latency. The main goal of optimizing the path is to ensure that data can be transmitted to the central data processing module with the lowest latency, the highest bandwidth, and the lowest packet loss rate, so as to provide reliable data support for subsequent analysis and decision-making. The specific steps are as follows:

[0029] S201: Set a maximum allowable latency threshold for each communication path. When the latency of a certain communication path exceeds this range, the path switching mechanism will be triggered. At the same time, priority scheduling is implemented, and the data with high priority in the data packet will be preferentially transmitted through the low-latency path, and the data with low priority in the data packet will be transmitted through the higher-latency path.

[0030] S202: Regularly monitor the bandwidth of each communication path to ensure that the latest bandwidth data can be obtained during data transmission, and automatically switch to the path with a wider bandwidth based on the monitoring results to ensure smooth data transmission.

[0031] S203: Regularly detect the transmission success rate of each communication path. If packet loss or transmission error is detected, immediately switch to the backup path and re-transmit the lost data packet through this path.

[0032] By adopting the above technical solution, based on the dynamic selection of the data transmission path, it can be ensured that no matter how the network conditions change, the data can be transmitted through the optimal path. Especially in the case of tight bandwidth or poor network quality, the path optimization can reduce latency and improve data transmission efficiency. At the same time, it can perform real-time monitoring and dynamic path adjustment according to information such as the bandwidth, latency, and packet loss rate of the communication path to ensure that high-priority data can be transmitted to the central data processing module in real time, providing timely and accurate data analysis support, which enables the equipment monitoring to respond quickly and discover potential problems as early as possible. The path optimization can effectively reduce the probability of data loss, and for the high-priority data necessary for real-time monitoring and control, ensure that the data is transmitted to the central data processing module smoothly and without error.

[0033] A further setting is that in the step S300, the process of data preprocessing includes:

[0034] S301. Adopt a dynamic threshold detection mechanism to clean abnormal data, dynamically adjust the threshold in combination with the change trend of the historical data of each sensor, and automatically adjust the threshold range after each data sampling period, so as to adapt to the load changes and environmental fluctuations of different environmental protection equipment;

[0035] During the operation of each sensor, dynamically adjust the static threshold of the corresponding sensor by calculating the moving average and standard deviation of the collected corresponding data;

[0036]

[0037] where x t is the sensor data at the current moment, n is the window size, and μ(t) and σ(t) respectively represent the data mean and standard deviation within the window;

[0038] Then, dynamically adjust the threshold range through the following formula:

[0039] Lower_Bound(t) = μ(t) - α·σ(t)

[0040] Upper_Bound(t) = μ(t) + α·σ(t)

[0041] where α is a dynamic adjustment coefficient to ensure that the threshold can adapt to changes when the data fluctuation increases;

[0042] Finally, perform the cleaning of abnormal data. When a certain data x t exceeds the threshold range adjusted by step S301, that is, x t < Lower_Bound(t) or x t > Upper_Bound(t), then the data is considered abnormal and is deleted or marked as invalid data;

[0043] Step S302. Adopt a wavelet denoising algorithm to remove noise;

[0044] Step S303. To avoid the existence of duplicate data and ensure that each data is only recorded once, adopt a deduplication method combining a time window and a hash algorithm;

[0045] First, each data packet is marked with a timestamp, and all received data packets are sorted according to the timestamp; within a set time window, if the contents of multiple data packets are the same, then these data packets are considered duplicates, and the older data packets are deleted, and only the latest one is retained;

[0046] Secondly, generate a hash value for the content of each data packet, mapping the data packet content to a unique hash code; when a new data packet arrives, calculate the hash value and compare it with the stored hash values; if the same hash value already exists, consider the data packet as duplicate data and perform deduplication.

[0047] By adopting the above technical solutions, the dynamic threshold detection mechanism can dynamically adjust the effective range of data according to the changing trend of historical data, thereby effectively filtering out abnormal, invalid, or incorrect data, which ensures that the data entering the central data processing module has high quality, thus improving the reliability of the analysis results; through the wavelet denoising algorithm, the system can effectively remove noise in complex signals, especially in physical signals such as vibration and pressure. The removal of noise can improve the accuracy of data and the effectiveness of analysis; through the deduplication method combining the hash algorithm and the time window, it can quickly and accurately identify and eliminate duplicate data, reduce the processing pressure of redundant data, and improve the operating efficiency of the system.

[0048] A further setting is that in the step S400, specifically:

[0049] S401. Standardize and normalize the data of each sensor to ensure that data with different dimensions can be processed on the same platform;

[0050] S402. Implement the analysis of cross-sensor data through a multi-modal collaborative analysis model to output a fault prediction report to help management personnel timely understand the operating status and potential risks of the equipment.

[0051] By adopting the above technical solutions, by standardizing and normalizing the data of each sensor, the system can ensure that data with different units and dimensions from different sensors can be processed consistently on the same platform, which greatly simplifies the data fusion and analysis, enabling seamless integration of multiple sensor data; the standardized and normalized data can be better fused across sensors, thereby realizing cross-sensor data interaction, and identifying false alarms caused by single sensor failures through cross-validation of multi-sensor data, improving the judgment accuracy.

[0052] A further setting is that in the step S402, specifically:

[0053] Step S4021. Assign a dynamic weight W i (t) to the data of each sensor, and its calculation method is:

[0054]

[0055] where C iC(t) is the credibility of sensor data, which is obtained from the data quality assessment after cleaning abnormal data in step S301. The calculation formula is C i (t) = 1 - (the amount of recent abnormal data / the total amount of data); S i S(t) is the sensor importance coefficient;

[0056] Step S402: Input the weighted multi-sensor data into the adaptive principal component analysis module, dynamically adjust the number of principal components to retain more than 95% of the information volume. The fused feature vector F(t) is expressed as:

[0057]

[0058] where X i norm (t) is the sensor data after standardization and normalization in step S401;

[0059] Step S403: Perform a hard rule check on the key parameters in the fused feature vector F(t). If the rule is triggered, give priority to outputting the rule judgment result and freeze the prediction result of the multi-modal collaborative analysis model until the parameters return to the normal range; when the rule is not triggered, execute step S404; and select the optimal model in the multi-modal collaborative analysis model;

[0060] Step S404:

[0061] If the data fluctuation is gentle and the load is stable, enable the lightweight time series prediction model to predict the device status in the next 1 hour;

[0062] If multi-parameter correlation anomalies are detected, activate the graph neural network to analyze the fault propagation path between the internal components of the environmental protection equipment;

[0063] When the external environmental variables mutate, call the reinforcement learning model to simulate the impact of different regulation strategies on the energy efficiency of the environmental protection equipment and generate the optimal control parameters.

[0064] By adopting the above technical solutions, each sensor data is assigned a dynamic weight according to its credibility and importance coefficient. This dynamic weighting mechanism can adjust the influence of sensors according to the actual operating conditions of the device, the working mode, and the current state of the sensors. When the device fails or the working load changes, important sensors can automatically obtain higher weights, thereby enhancing their influence in prediction. This enables the system to optimize the accuracy and reliability of prediction according to different application scenarios. By inputting the weighted multi-sensor data into the adaptive principal component analysis module, the data collected by various sensors can be fully utilized and analyzed from different angles and dimensions. This fusion can greatly improve the comprehensiveness and integrity of the data, helping the system to better understand and monitor the overall operating conditions of the device. Especially for the complex relationship between environmental parameters and device states, data fusion provides a more accurate basis for subsequent fault prediction. The fused feature vector F(t) can provide higher-quality input for subsequent analysis and prediction.

[0065] A further setting is that in the step S404, the generated optimal control parameter is set to P SET , and this optimal control parameter P SET needs to be compared with the physical limit of the environmental protection device to ensure that it satisfies:

[0066] P SET ∈[P min , P max and dP SET / dt ≤ ΔPthreshold

[0067] where P min and P max are respectively the minimum and maximum controls allowed for the environmental protection device, dP SET / dt is the time change rate of the control parameter, and ΔPthreshold is the parameter change rate threshold to prevent sudden impact.

[0068] By adopting the above technical solutions, by comparing the control parameter P SET with the physical limits P min and P max of the device, it is ensured that the control parameter always remains within the safe range allowed by the device. This physical safety check can effectively avoid the risks of over-regulation or overloading of the device, and reduce equipment damage or failures caused by improper setting of control parameters. This is particularly important for high-load and high-frequency operating environmental protection devices, and can ensure the stable operation of the device under any circumstances. By introducing the parameter change rate threshold ΔPthreshold, the adjustment range of the control parameter is effectively limited, avoiding unstable device operation caused by too fast or sudden parameter changes, which has a positive effect on the long-term stability of the device.

[0069] In summary, the present invention has the following beneficial effects:

[0070] Through intelligent, dynamically adaptive, and data-driven real-time decision support, it can improve the monitoring accuracy of environmental protection equipment, optimize the acquisition strategy, enhance the fault prediction ability, and significantly reduce the downtime and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a schematic flowchart of an embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0072] The present invention will be further described in detail below with reference to the accompanying drawings.

[0073] As shown in the Figure 1 accompanying drawings;

[0074] This embodiment discloses an environmental protection equipment management and data analysis system based on the Internet of Things, including an Internet of Things sensor group installed on each environmental protection equipment, a central data processing module that realizes data intercommunication with the Internet of Things sensor groups on each environmental protection equipment through Internet of Things communication, and a cloud database that can receive the data of the central data processing module; the specific process of the system operation includes the following steps:

[0075] S100. Comprehensively collect data through the Internet of Things sensor groups on each environmental protection equipment. The data includes the working parameters of each environmental protection equipment and its external environment data. The Internet of Things sensor group includes multiple different types of sensors; the Internet of Things sensor group includes, but is not limited to, temperature sensors, humidity sensors, pressure sensors, gas sensors, vibration sensors, flow sensors, electric energy metering sensors, dissolved oxygen sensors, CO2 concentration sensors, ammonia nitrogen sensors, soot sensors, and ultraviolet sensors; the Internet of Things sensor group can not only collect data during the normal operation of each environmental protection equipment, but also send warning signals when the equipment fails or operates abnormally; different types of environmental protection equipment can carry the required sensors in the Internet of Things sensor group according to the actual working conditions to ensure that the monitoring requirements of all key parameters are covered;

[0076] S200. Each Internet of Things sensor group installed on the environmental protection equipment transmits the collected data to the central data processing module through the Internet of Things communication network;

[0077] S300. After the central data processing module receives the data transmitted by each Internet of Things sensor, it preprocesses the data to ensure that the received data is accurate and reliable;

[0078] S400. Perform multi-dimensional analysis and calculation on the preprocessed data to output a fault prediction report to help managers understand the operating status and potential risks of environmental protection equipment in a timely manner;

[0079] S500. Upload all the data from steps S100 to S400 to the cloud database for storage and management; the cloud database supports multi-level data backup and redundancy mechanisms, which can ensure the durability and integrity of data to the greatest extent; in addition, the cloud database also has efficient data query and analysis capabilities, which can quickly obtain the required information from a large amount of data, support managers to monitor device data in real time, query historical records, and make intelligent decisions based on the data.

[0080] Among them, in step S100, each sensor will collect data regularly according to the preset sampling period, and the sampling period is dynamically adjusted based on the monitoring requirements. During the dynamic adjustment process, the sampling frequency will be automatically adjusted through an adaptive algorithm according to the load status, battery energy efficiency, and external environmental variable factors of the environmental protection equipment to balance the real-time performance and energy efficiency of data collection.

[0081] The performance of the adaptive algorithm for automatically adjusting the sampling frequency in this embodiment is as follows: when the environmental protection equipment is in a high-load state, the sampling frequency will be automatically increased to ensure more accurate monitoring of the operation status of the environmental protection equipment; for example, the environmental protection equipment may require higher-frequency temperature and current sampling during peak load periods to monitor whether it is overloaded. When the environmental protection equipment is powered by a battery, the sampling frequency will be dynamically adjusted according to the remaining battery power; when the battery power is low, the sampling frequency will be reduced to extend the battery life. When environmental factors such as temperature, humidity, and gas concentration change significantly, the sampling frequency will be automatically increased to capture these changes; for example, when the gas concentration changes significantly, increasing the sampling frequency helps to give an early warning of possible pollution or danger in a timely manner.

[0082] Among them, the specific steps of the adaptive algorithm for automatically adjusting the sampling frequency are as follows:

[0083] S101. Make a first adjustment to the sampling frequency based on the load status of the environmental protection equipment:

[0084] f L (t) = f0 × (1 + α·L(t))

[0085] Among them, f L (t) is the influence of the equipment load on the sampling frequency, f0 is the default sampling frequency, α is the load adjustment factor, L(t) is the current load status of the environmental protection equipment, and t is the timestamp; the higher L(t) is, the higher the sampling frequency f L (t) is.

[0086] S102. Make a second adjustment to the sampling frequency based on the battery energy efficiency of the environmental protection equipment under the influence of the load status of the environmental protection equipment:

[0087] f ηf(t) = f(t) × (1 - β·η(t)) L × (1 - β·η(t))

[0088] where f(t) is the influence of the battery energy efficiency of the environmental protection equipment on the sampling frequency, β is the energy efficiency adjustment factor, and η(t) is the current battery energy consumption of the environmental protection equipment; when η(t) is low, the sampling frequency f(t) will decrease; η f(t) is the influence of the battery energy efficiency of the environmental protection equipment on the sampling frequency, β is the energy efficiency adjustment factor, and η(t) is the current battery energy consumption of the environmental protection equipment; when η(t) is low, the sampling frequency f(t) will decrease; η will decrease;

[0089] S103. Under the influence of the battery energy efficiency of the environmental protection equipment, the sampling frequency is adjusted three times based on the external environmental variables:

[0090] f(t) = f(t) × (1 + γ·E(t)) E f(t) = f(t) × (1 + γ·E(t)) η × (1 + γ·E(t))

[0091] where f(t) is the influence of the external environmental variables on the sampling frequency, γ is the environmental adjustment factor, and E(t) is the current external environmental variable; when E(t) changes greatly, the sampling frequency will increase; E f(t) is the influence of the external environmental variables on the sampling frequency, γ is the environmental adjustment factor, and E(t) is the current external environmental variable; when E(t) changes greatly, the sampling frequency will increase;

[0092] S104. After adjusting the sampling frequency three times, according to the weighted average of the load status, battery energy efficiency, and external environmental variable factors of the environmental protection equipment, the dynamic sampling frequency f(t) of the environmental protection equipment can be obtained as follows: dynamic is:

[0093] f(t) = w1·f(t) + w2·f(t) + w3·f(t) dynamic f(t) = w1·f(t) + w2·f(t) + w3·f(t) L + w2·f(t) η + w3·f(t) E + w3·f(t)

[0094] where w1, w2, and w3 are the weight coefficients of the load status, battery energy efficiency, and external environmental variables of the environmental protection equipment respectively.

[0095] The significance of the weight assignment in step S104 in this embodiment is as follows:

[0096] The influences of the load status, battery energy efficiency, and external environmental variable factors of the environmental protection equipment on the sampling frequency may be independent, but they may also be intertwined with each other. When the equipment load is low, the influence of energy efficiency may be greater; while when the equipment load is high, the influence of the load itself may be more significant. Therefore, it is necessary to dynamically balance the actual influence of each factor on the sampling frequency by adjusting the weights. Adjusting the weights of each factor can make the system more flexible and automatically adapt to different equipment usage scenarios and operating states.

[0097] Furthermore, under different environmental and device load conditions, the influence of the load status, battery energy efficiency, and external environmental variable factors of environmental protection equipment on the sampling frequency may vary. In a scenario where energy conservation is prioritized, energy efficiency needs to be given a higher weight; while in a scenario of a device with a higher load, the impact of the load is more important. Therefore, by dynamically adjusting the weights, different application requirements can be met according to the actual situation.

[0098] Among them, in step S200, it is necessary to select the best network path in the Internet of Things communication network to improve data transmission efficiency and reduce latency. The main goal of optimizing the path is to ensure that data can be transmitted to the central data processing module with the lowest latency, the highest bandwidth, and the lowest packet loss rate, thereby providing reliable data support for subsequent analysis and decision-making. The specific steps are as follows:

[0099] S201. Set a maximum allowable delay threshold for each communication path. When the delay of a certain communication path exceeds this range, the path switching mechanism will be triggered; at the same time, priority scheduling is implemented, and data with a higher priority in the data packet will be preferentially transmitted through a low-latency path, while data with a lower priority in the data packet will be transmitted through a higher-latency path.

[0100] S202. Regularly monitor the bandwidth of each communication path to ensure that the latest bandwidth data can be obtained during data transmission, and automatically switch to a path with a wider bandwidth based on the monitoring results to ensure smooth data transmission.

[0101] S203. Regularly detect the transmission success rate of each communication path. If packet loss or transmission error is detected, immediately switch to the backup path and re-transmit the lost data packets through this path.

[0102] Among them, in step S300, the process of data preprocessing includes:

[0103] In the Internet of Things system, the data collected by sensors often contains noise, duplicate, and abnormal data. Effective data preprocessing can not only improve data quality but also provide accurate input for subsequent analysis.

[0104] S301. In traditional data cleaning, static thresholds are usually used to filter out abnormal data based on pre-set maximum and minimum values. However, this method fails to effectively adapt to the changes of different devices. Therefore, in this embodiment, a dynamic threshold detection mechanism is adopted to clean abnormal data, dynamically adjust the threshold in combination with the change trend of the historical data of each sensor, and automatically adjust the threshold range after each data sampling period, so as to adapt to the load changes and environmental fluctuations of different environmental protection equipment.

[0105] During the operation of each sensor, the static threshold of the corresponding sensor is dynamically adjusted by calculating the moving average and standard deviation of the collected corresponding data; when the load is high, the environmental protection equipment generates more heat, resulting in a slight increase in temperature; at this time, by adjusting the threshold of the temperature sensor, false alarms triggered by load changes can be avoided. If the humidity in the environment where the environmental protection equipment is located changes drastically, the originally set static threshold may not be able to reflect these changes in a timely manner. By dynamically adjusting the threshold of the humidity sensor, the accuracy of data processing can be ensured: the calculation formulas for the data mean and standard deviation of each sensor are as follows:

[0106]

[0107] where x t is the sensor data at the current moment, n is the window size, and μ(t) and σ(t) represent the data mean and standard deviation within the window respectively;

[0108] Next, the threshold range is dynamically adjusted by the following formula:

[0109] Lower_Bound(t) = μ(t) - α·σ(t)

[0110] Upper_Bound(t) = μ(t) + α·σ(t)

[0111] where α is the dynamic adjustment coefficient to ensure that the threshold can adapt to changes when the data fluctuation increases;

[0112] Finally, for the cleaning of abnormal data, when a certain data x t exceeds the threshold range adjusted in step S301, that is, x t < Lower_Bound(t) or x t > Upper_Bound(t), the data is considered abnormal and is either excluded or marked as invalid data;

[0113] Step S302: Use the wavelet denoising algorithm to remove noise; in complex physical signals, such as pressure and vibration signals, wavelet transform can better capture the local characteristics of the signal and can better adapt to multiple different types of sensors.

[0114] Step S303: To avoid the existence of duplicate data and ensure that each data is recorded only once, a deduplication method combining a time window and a hash algorithm is used;

[0115] First, each data packet is marked with a timestamp, and all received data packets are sorted according to the timestamp; within a set time window, if the contents of multiple data packets are the same, specifically, the device ID, sensor type, and timestamp are all the same, then these data packets are considered duplicates, and the older data packets are deleted, only retaining the latest one;

[0116] Secondly, a hash value is generated for the content of each data packet to map the data packet content to a unique hash code; when a new data packet arrives, the hash value is calculated and compared with the stored hash values; if the same hash value already exists, then the data packet is considered duplicate data and deduplication is performed.

[0117] After step S303, duplicate data can be judged and eliminated efficiently and accurately. The hash value is used to quickly detect absolutely duplicate data. This deduplication method can ensure high efficiency when processing a large amount of sensor data while taking into account the accuracy of calculations.

[0118] Among them, specifically in step S400:

[0119] S401. Standardize and normalize the data of each sensor to ensure that data with different dimensions can be processed on the same platform;

[0120] S402. Implement cross-sensor data analysis through a multi-modal collaborative analysis model to output a fault prediction report to help managers understand the operating status and potential risks of the equipment in a timely manner.

[0121] Among them, specifically in step S402:

[0122] Step S4021. Assign a dynamic weight W i (t) to the data of each sensor, and its calculation method is:

[0123]

[0124] Among them, C i (t) is the credibility of the sensor data, which is obtained from the data quality assessment after cleaning abnormal data through step S301. The calculation formula is C i (t) = 1 - (recent abnormal data volume / total data volume); S i (t) is the sensor importance coefficient; the numerator C i (t)·S i (t) represents the product of the credibility and importance of the current sensor i; the denominator represents the sum of the products of all sensors and is used to normalize the weight.

[0125] Step S402: Input the weighted multi-sensor data into the adaptive principal component analysis module, dynamically adjust the number of principal components to retain more than 95% of the information volume, and the fused feature vector F(t) is expressed as:

[0126]

[0127] where X i norm (t) is the sensor data standardized and normalized in Step S401;

[0128] Step S403: Perform a hard rule check on the key parameters in the fused feature vector F(t), such as "if the voltage fluctuation exceeds ±10%, trigger an emergency shutdown suggestion". If the rule is triggered, give priority to outputting the rule judgment result and freeze the prediction result of the multi-modal collaborative analysis model until the parameters return to the normal range; when the rule is not triggered, execute Step S404; and select the optimal model in the multi-modal collaborative analysis model;

[0129] Step S404:

[0130] If the data fluctuation is gentle and the load is stable, enable the lightweight time series prediction model to predict the device status in the next 1 hour; this model is based on the time series trend of the data, combines the laws of historical data, predicts the possible health status and operation problems of the device in the short term, and gives early warnings of possible device failures;

[0131] If multi-parameter correlation anomalies are detected, activate the graph neural network to analyze the fault propagation path between the internal components of the environmental protection equipment;

[0132] When the external environmental variables mutate, call the reinforcement learning model to simulate the impact of different regulation strategies on the energy efficiency of the environmental protection equipment and generate the optimal control parameters. This process ensures that the equipment can maintain efficient operation in a complex external environment and minimize the occurrence of failures.

[0133] where, in Step S404, the generated optimal control parameter is set as P SET , and this optimal control parameter P SET needs to be compared with the physical limit of the environmental protection equipment to ensure that it satisfies:

[0134] P SET ∈[P min ,P max and dP SET / dt ≤ ΔPthreshold

[0135] where P min and P max are the minimum and maximum controls allowed for the environmental protection equipment respectively, and dP SET / dt is the time change rate of the control parameter, and ΔPthreshold is the parameter change rate threshold to prevent sudden changes and impacts.

[0136] Assume that the change in the fan speed is required not to exceed 500 revolutions per minute, that is, the change rate per second is 500 revolutions / 60 seconds. Then it is necessary to satisfy dP SET / dt ≤ 500 revolutions / 60 seconds. This condition ensures that when adjusting the control parameter, the fan speed will not change suddenly and always remains within the physical limits of the equipment, thus guaranteeing the stability and safety of the equipment operation.

[0137] Example 1

[0138] In the bag dust removal system supporting the No. 3 blast furnace of a certain iron and steel plant, the equipment model is LDM - 1000, the rated power is 75 kW, the processing air volume is 100,000 m 3 / h, and the filtration area is 1200 m 2 , and the working temperature range is 60 - 80 °C.

[0139] In step S100, the Internet of Things sensor group installed on the bag dust removal system includes a temperature sensor, a pressure sensor, a dust sensor, a vibration sensor, and an electric energy metering sensor; among them:

[0140] The temperature sensor model is PT100 and is installed in the inlet pipe of the dust collector;

[0141] The pressure sensor model is Honeywell 26PC and is installed in the bag filter bin;

[0142] The dust sensor model is TSI 8530 and is installed at the discharge port;

[0143] The vibration sensor model is Wilcoxon 786 and is installed on the fan bearing seat;

[0144] The electric energy metering sensor model is ADE7953.

[0145] After the adaptive algorithm automatically adjusts the sampling frequency for global configuration in step S100, the actual situation is as follows:

[0146] S101. Make a first adjustment to the sampling frequency based on the load status of the environmental protection equipment:

[0147] The default sampling frequency f0 = 1 time / minute;

[0148] When the load rate L(t) increases from 50% to 90%, the adjustment factor α = 0.25, and the frequency is increased to f L = 1 + 0.25×(0.9 - 0.5) = 1.1 times / minute.

[0149] S102. Under the influence of the load state of the environmental protection equipment, the sampling frequency is adjusted twice based on the battery energy efficiency of the environmental protection equipment:

[0150] The battery is a lithium battery pack. When the power of the lithium battery pack is lower than 20%, the adjustment factor β = 0.4, and the sampling frequency drops to f η = 1.1×(1 - 0.4×0.8) = 0.748 times / minute.

[0151] S103. Under the influence of the battery energy efficiency of the environmental protection equipment, the sampling frequency is adjusted three times based on external environmental variables:

[0152] When the temperature fluctuation ΔT of the inlet pipe of the dust collector > 10°C / hour, the adjustment factor γ = 0.2, and the frequency is increased to f E = 0.748×(1 + 0.2) = 0.898 times / minute.

[0153] The weight distribution in step S104 is w1 = 0.5, w2 = 0.3, w3 = 0.2, and the final dynamic frequency is f dynamic = 0.5×1.1 + 0.3×0.748 + 0.2×0.898 = 0.95 times / minute.

[0154] Taking the threshold adjustment of the temperature sensor in step S300 as an example:

[0155] The window size n in step S301 = 60 minutes, the mean value μ(t) = 68°C, and the standard deviation σ(t) = 4°C;

[0156] In step S302:

[0157] Lower Bound = 68 - 1.5×4 = 62°C, Upper Bound = 68 + 1.5×4 = 74°C.

[0158] If the sampled temperature suddenly rises to 78°C (4°C above the threshold), an abnormal mark is triggered, and in combination with the pressure data (rising from 4.2 kPa to 6.8 kPa), it is determined as "bag clogging".

[0159] The final feedback is as follows:

[0160] 1. If the pressure > 6 kPa and the temperature > 75°C for 3 consecutive times, the system automatically sends an "emergency shutdown" instruction and starts the standby dust removal unit;

[0161] 2. The bag replacement cycle is extended from 3 months to 5 months, and the annual maintenance cost is reduced by 28%.

[0162] Example 2

[0163] In the sewage treatment system of a certain sewage treatment plant, 5 variable-frequency blowers are equipped. The model of the variable-frequency blower is RHG-300, with a single-machine power of 30 kW and an air volume of 0 - 3000 m 3 / h.

[0164] In step S100, the Internet of Things sensor group installed in the sewage treatment system includes a dissolved oxygen sensor, an ammonia nitrogen sensor, a flow sensor, and an electric energy metering sensor; among them:

[0165] The model of the dissolved oxygen sensor is Hach LDO, and it is installed in the aerobic tank;

[0166] The model of the ammonia nitrogen sensor is Endress+Hauser CA80AM, the model of the flow sensor is Krohne OPTIMASS6400, and the electric energy metering sensor uses the electric energy quality recorder with the model of Fluke 1738.

[0167] In step S200, the main path of the Internet of Things communication network is NB-IoT, and the backup path is LoRaWAN;

[0168] When the NB-IoT delay > 200 ms or the packet loss rate > 3%, switch to LoRaWAN

[0169] The priority scheduling is as follows: The real-time data of dissolved oxygen is preferentially transmitted, and the historical data is uploaded in batches.

[0170] In step S402, the credibility C of the dissolved oxygen sensor data i (t) = 0.95 (recent abnormal rate 5%), the importance coefficient S of the dissolved oxygen sensor i (t) = 0.9; After calculation, the weight W of the dissolved oxygen sensor i (t) = 0.58;

[0171] The credibility C of the flow sensor data i (t) = 0.98, the importance coefficient S of the flow sensor i (t) = 0.7; After calculation, the weight W of the flow sensor i (t) = 0.42.

[0172] In step S404, when the influent ammonia nitrogen concentration > 30 mg / L and the dissolved oxygen < 1.5 mg / L, the rotational speed is increased to 2000 RPM, and the aeration volume is increased by 40%; when the influent ammonia nitrogen < 15 mg / L and the dissolved oxygen > 2.5 mg / L, the rotational speed is decreased to 1200 RPM, and the aeration volume is decreased by 30%; Through precise aeration control, the daily average energy consumption is reduced from 450 kWh to 320 kWh, saving 28.9%.

[0173] The final feedback is as follows:

[0174] The average daily operation duration of a single variable-frequency blower is reduced from 18 hours to 14 hours, and the monthly power savings is 30 kW × 4 h × 30 days = 3600 kW.

[0175] This specific embodiment is only an explanation of the present invention and is not a limitation thereof. After reading this specification, those skilled in the art can make modifications to this embodiment without creative contributions as needed, but as long as it is within the scope of the claims of the present invention, it is protected by the patent law.

Claims

1. An environmental protection equipment management and data analysis system based on the Internet of Things, comprising an Internet of Things sensor group installed on each environmental protection equipment, a central data processing module that realizes data intercommunication with the Internet of Things sensor groups on each environmental protection equipment through Internet of Things communication, and a cloud database capable of receiving data from the central data processing module; characterized in that, The specific process of system operation includes the following steps: S100. Comprehensively collect data through the Internet of Things sensor groups on each environmental protection device. The Internet of Things sensor groups include multiple different types of sensors; S200. Each Internet of Things sensor group installed on the environmental protection device transmits the collected data to the central data processing module through the Internet of Things communication network; S300. After the central data processing module receives the data transmitted by each Internet of Things sensor, it preprocesses the data; S400. Perform multi-dimensional analysis and calculation on the preprocessed data to output a fault prediction report to help managers understand the operation status and potential risks of environmental protection devices in a timely manner; S500. Upload all the data from step S100 to step S400 to the cloud database for storage and management.

2. An environmental protection equipment management and data analysis system based on the Internet of Things according to claim 1, characterized in that: In step S100, each sensor collects data at regular intervals according to a preset sampling period. The sampling period is dynamically adjusted based on monitoring requirements. During the dynamic adjustment process, the sampling frequency is automatically adjusted through an adaptive algorithm according to the load status, battery energy efficiency, and external environmental variable factors of the environmental protection device to balance the real-time performance and energy efficiency of data collection.

3. An environmental protection equipment management and data analysis system based on the Internet of Things according to claim 2, characterized in that: The specific steps for the adaptive algorithm to automatically adjust the sampling frequency are as follows: S101. Make a first adjustment to the sampling frequency based on the load status of the environmental protection device: f L f(t) = f0 × (1 + α·L(t)) Among them, f L (t) is the influence of the device load on the sampling frequency, f0 is the default sampling frequency, α is the load adjustment factor, L(t) is the current load state of the environmental protection device, and t is the timestamp; S102. Under the influence of the load status of the environmental protection device, make a second adjustment to the sampling frequency based on the battery energy efficiency of the environmental protection device: f η f(t) = L f(t) × (1 - β·η(t)) Among them, f η (t) is the influence of the battery energy efficiency of the environmental protection equipment on the sampling frequency, β is the energy efficiency adjustment factor, and η(t) is the current battery energy consumption of the environmental protection equipment; S103. Under the influence of the battery energy efficiency of the environmental protection device, make a third adjustment to the sampling frequency based on external environmental variables: f E f(t) = η f(t) × (1 + γ · E(t)) where, f E (t) is the influence of the external environmental variable on the sampling frequency, γ is the environmental adjustment factor, and E(t) is the current external environmental variable; S104. After adjusting the sampling frequency three times, the dynamic sampling frequency f of the environmental protection device can be obtained according to the weighted average of the load status of the environmental protection device, the battery energy efficiency, and the external environmental variable factors. dynamic (t) is as follows: f dynamic f(t) = w1·f L (t) + w2·f η (t) + w3·f E (t) Among them, w1, w2, and w3 are the weight coefficients of the load status, battery energy efficiency, and external environmental variables of the environmental protection device respectively.

4. An environmental protection equipment management and data analysis system based on the Internet of Things according to claim 1, characterized in that, In step S200, it is necessary to select the best network path in the Internet of Things communication network to improve data transmission efficiency and reduce latency. The main goal of optimizing the path is to ensure that data can be transmitted to the central data processing module with the lowest latency, the highest bandwidth, and the lowest packet loss rate, so as to provide reliable data support for subsequent analysis and decision-making. The specific steps are as follows: S201. Set a maximum allowable delay threshold for each communication path. When the delay of a certain communication path exceeds this range, a path switching mechanism will be triggered; at the same time, priority scheduling is implemented, and the data with a high priority in the data packet will be preferentially transmitted through the low-latency path, and the data with a low priority in the data packet will be transmitted through the higher-latency path; S202. Regularly monitor the bandwidth of each communication path to ensure that the latest bandwidth data can be obtained during data transmission, and automatically switch to the path with a wider bandwidth based on the monitoring results to ensure smooth data transmission; S203. Regularly detect the transmission success rate of each communication path. If packet loss or transmission error is detected, immediately switch to the backup path and re-transmit the lost data packets through this path.

5. An environmental protection equipment management and data analysis system based on the Internet of Things according to claim 1, characterized in that, In step S300, the process of data preprocessing includes: S301. Adopt a dynamic threshold detection mechanism to clean abnormal data, dynamically adjust the threshold in combination with the change trend of the historical data of each sensor, and automatically adjust the threshold range after each data sampling period, so as to adapt to the load changes and environmental fluctuations of different environmental protection equipment; During the operation of each sensor, dynamically adjust the static threshold of the corresponding sensor by calculating the moving average and standard deviation of the collected corresponding data; where x t is the sensor data at the current moment, n is the window size, and μ(t) and σ(t) represent the mean and standard deviation of the data within the window, respectively; Next, dynamically adjust the threshold range through the following formula: Lower_Bound(t) = μ(t) - α·σ(t) Upper_Bound(t) = μ(t) + α·σ(t) where α is a dynamic adjustment coefficient to ensure that the threshold can adapt to changes when the data fluctuation increases; Finally, perform cleaning on the abnormal data. When a certain data x t exceeds the threshold range adjusted in step S301, that is, x t <Lower_Bound(t) or x t >Upper_Bound(t), then the data is considered abnormal and is removed or marked as invalid data; Step S302. Adopt a wavelet denoising algorithm to remove noise; Step S303. In order to avoid the existence of duplicate data and ensure that each data is only recorded once, adopt a deduplication method combining a time window and a hash algorithm; First, each data packet is marked with a timestamp, and all received data packets are sorted according to the timestamp; within a set time window, if the contents of multiple data packets are the same, these data packets are considered duplicates, and the older data packets are deleted, only keeping the latest one; Secondly, generate a hash value for the content of each data packet, map the data packet content to a unique hash code; when a new data packet arrives, calculate the hash value and compare it with the stored hash value; if the same hash value already exists, the data packet is considered duplicate data and deduplicated.

6. An environmental protection equipment management and data analysis system based on the Internet of Things according to claim 5, characterized in that, Specifically in the step S400 described above: S401. Standardize and normalize the data of each sensor to ensure that data with different dimensions can be processed on the same platform; S402. Implement cross-sensor data analysis through a multi-modal collaborative analysis model to output a fault prediction report to help managers timely understand the operating status and potential risks of the equipment.

7. An environmental protection equipment management and data analysis system based on the Internet of Things according to claim 6, characterized in that, Specifically in the step S402 described above: Step S4021: Assign a dynamic weight W i (t) to the data of each sensor, and its calculation method is as follows: Among them, C i (t) is the reliability of sensor data, which is obtained from the data quality assessment after cleaning abnormal data through step S301. The calculation formula is C i (t) = 1 - (the amount of recent abnormal data / the total amount of data); S i (t) is the importance coefficient of the sensor; Step S402. Input the weighted multi-sensor data into an adaptive principal component analysis module, dynamically adjust the number of principal components to retain more than 95% of the information volume, and the fused feature vector F(t) is expressed as: Among them, X i norm (t) is the sensor data after being standardized and normalized in step S401; Step S403. Perform a hard rule check on the key parameters in the fused feature vector F(t). If the rule is triggered, give priority to outputting the rule judgment result and freeze the prediction result of the multi-modal collaborative analysis model until the parameters return to the normal range; when the rule is not triggered, execute step S404; and select the optimal model in the multi-modal collaborative analysis model; Step S404. If the data fluctuation is gentle and the load is stable, enable a lightweight time series prediction model to predict the equipment status in the next 1 hour; If multi-parameter correlation anomalies are detected, activate the graph neural network to analyze the fault propagation path between the internal components of the environmental protection equipment; When the external environmental variables mutate, call the reinforcement learning model to simulate the impact of different control strategies on the energy efficiency of the environmental protection equipment and generate optimal control parameters.

8. An environmental protection equipment management and data analysis system based on the Internet of Things according to claim 7, characterized in that, In the step S404, the generated optimal control parameter is set to P SET , and the optimal control parameter P SET needs to be compared with the physical limit of the environmental protection equipment to ensure that it meets: P SET ∈ [P min , P max and dP SET / dt ≤ ΔPthreshold where P min and P max are the minimum and maximum controls allowed for the environmental protection equipment respectively, dP SET / dt is the time change rate of the control parameter, and ΔPthreshold is the parameter change rate threshold to prevent sudden impact.

Citation Information

Cited By

  • Internet of Things data flow intelligent management system based on data processing

    CN120935224A

  • An internet of things data stream intelligent management system based on data processing

    CN120935224B

  • Public transport wireless connection quality real-time monitoring system and method

    CN122420907A