Environmental protection equipment monitoring and regulation system based on sensor network

Abstract

The invention relates to the technical field of data processing systems, in particular to an environmental protection equipment monitoring and regulation system based on a sensor network, which comprises an environmental reference model construction module, a low-power-consumption inspection scheduling module, a sudden deviation detection module and an emergency cooperative regulation module. According to the method, a periodic environment reference model is constructed, a stable fluctuation upper bound is determined, an environment state is automatically identified and a working mode is adjusted by taking the stable fluctuation upper bound as a reference, a dormant state is switched to greatly reduce energy consumption when data is lower than a threshold value, and high-frequency sampling is triggered to accurately capture abnormity when a numerical value breaks through the upper bound; and after counting over-limit readings and secondary confirmation, automatic association is carried out and equipment full-power operation is activated, manual intervention is abandoned based on a data-driven hierarchical response mechanism, seamless connection from low-power-consumption inspection to emergency disposal is realized, long-term endurance of the equipment is ensured, and meanwhile, real-time and efficient blocking of an emergency pollution event is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing systems, and in particular to an environmental protection equipment monitoring and control system based on a sensor network. BACKGROUND

[0002] In the technical field of data processing systems, with the rapid development of technologies such as the Internet of Things, big data, and artificial intelligence, there is an increasing demand for intelligent monitoring and control of various devices and environments. Data processing systems mainly involve real-time processing, analysis, and storage of data collected by sensors, and making decisions or issuing control instructions based on the processing results to achieve state perception, event identification, fault diagnosis, and automated response of the physical world. Among them, the traditional environmental protection equipment monitoring and control system refers to collecting environmental protection equipment operation status and environmental parameter data by deploying various sensors or detectors. These data are usually aggregated to a central control unit through wired connection or simple wireless transmission, and then the device status and environmental changes are manually judged by the operator, and the device parameters are manually adjusted or the corresponding control strategy is started. For example, for wastewater treatment equipment, the traditional method may manually adjust the dosing amount or aeration intensity after manually sampling and analyzing water quality indicators; for exhaust emission equipment, it may manually operate the fan speed or start-stop of the filtering unit after detecting the emission concentration by fixed-position sensors.

[0003] Traditional environmental monitoring and control relies on manual experience judgment and manual operation execution. Sensors collect and transmit data at a fixed frequency, which is subjectively analyzed and decided by personnel. The lag response mechanism is difficult to capture instantaneous environmental fluctuations, and manual adjustment cannot accurately match real-time concentration changes, resulting in energy waste due to overrunning of equipment during low pollution periods, inability to timely suppress diffusion due to response delay during high pollution periods, lack of flexibility of single fixed monitoring mode, inability to achieve fine management in complex working conditions, increased operation and maintenance costs, and low treatment efficiency due to untimely issuance of instructions. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art and to provide an environmental protection equipment monitoring and control system based on a sensor network.

[0005] In order to achieve the above-mentioned purpose, the present application adopts the following technical solution: an environmental protection equipment monitoring and control system based on a sensor network comprises: An environmental benchmark model construction module collects pollutant concentration data of sensor nodes, calculates the mean value to form a daily mean benchmark curve and a standard deviation, superimposes three times the standard deviation on the daily mean benchmark curve to obtain a smooth fluctuation upper limit, combines the daily mean benchmark curve and the smooth fluctuation upper limit to generate a periodic environmental benchmark model, and transmits the model to a low-power inspection scheduling module and a sudden deviation detection module. A low-power inspection scheduling module calls the periodic environmental benchmark model to obtain the upper limit of the smooth fluctuation, compares the upper limit of the smooth fluctuation with a device start threshold, establishes a low-power sleep instruction if the upper limit of the smooth fluctuation is lower than the device start threshold, establishes a regular monitoring instruction if the upper limit of the smooth fluctuation is higher than the device start threshold, and synthesizes a sensor scheduling instruction; A sudden deviation detection module monitors real-time pollutant concentration readings, determines a potential deviation event if the real-time pollutant concentration readings exceed the upper limit of the smooth fluctuation provided by the periodic environmental benchmark model, generates a high-frequency sampling instruction, and transmits the high-frequency sampling instruction to an emergency collaborative control module; The emergency collaborative control module monitors high-frequency pollutant concentration readings based on the potential deviation event, checks a node-device association table if the number of high-frequency pollutant concentration readings exceeding the upper limit of the smooth fluctuation satisfies a confirmation threshold, and generates a full-power operation environmental protection device control instruction.

[0006] As a further scheme of the present application, the periodic environmental benchmark model includes a daily average value benchmark curve and an upper limit of smooth fluctuation, the sensor scheduling instruction includes a low-power sleep instruction and a regular monitoring instruction, the high-frequency sampling instruction includes a potential deviation event identifier and high-frequency pollutant concentration reading collection, and the full-power operation environmental protection device control instruction includes a device identifier, an operation mode, and a power level.

[0007] As a further scheme of the present application, the environmental benchmark model construction module has the following specific functions: A historical data cleaning submodule obtains pollutant concentration data of a sensor node in a preset historical period, eliminates outliers in the pollutant concentration data caused by sensor failure, extracts valid time series samples and aligns them according to timestamps, and generates a cleaned concentration data set; A statistical feature calculation submodule calculates an arithmetic mean value of all samples for each time point based on the cleaned concentration data set to construct a daily average value benchmark curve, calculates a standard deviation of sample distribution at the same time point, quantifies the fluctuation range of environmental background noise according to the normal distribution principle, and generates a time point statistical feature vector; A dynamic boundary synthesis submodule fuses the daily average value benchmark curve and the weighted standard deviation based on the time point statistical feature vector, introduces a historical extreme value fluctuation rate to adaptively correct the boundary, and uses a dynamic boundary synthesis formula: ; calculates an upper limit of smooth fluctuation at each time point, combines the daily average value benchmark curve and the upper limit of smooth fluctuation, and generates a periodic environmental benchmark model; wherein, represents the upper limit of smooth fluctuation at time t, represents a value on the daily average value benchmark curve at time t, representing a preset coefficient of three standard deviations, representing the standard deviation at time t, representing a global maximum value of the pollutant concentration data in the historical period, representing a global minimum value of the pollutant concentration data in the historical period, representing a global average value of the pollutant concentration data in the historical period.

[0008] As a further scheme of the present application, the specific functions of the low-power inspection scheduling module are implemented as: a safety margin analysis submodule, which analyzes the periodic environmental reference model, extracts a smooth fluctuation upper limit sequence in a future scheduling period, point-by-point calculates the difference between the smooth fluctuation upper limit sequence and the device start threshold, screens the smallest positive difference as a safety margin indicator, and generates a periodic safety evaluation report; a hibernation strategy formulation submodule, which, when the safety margin indicator is positive and greater than a preset safety buffer bandwidth, calculates a hibernation interval length in inverse proportion to the size of the safety margin indicator, plans a time node for the next wake-up, and constructs a low-power hibernation instruction; a monitoring task distribution submodule, which, when the safety margin indicator is less than or equal to the preset safety buffer bandwidth, sets a high-density regular sampling frequency, configures time slot parameters for data transmission, constructs a regular monitoring instruction, encapsulates the low-power hibernation instruction or the regular monitoring instruction, and generates a sensor scheduling instruction.

[0009] As a further scheme of the present application, the specific functions of the burst deviation detection module are implemented as: a real-time synchronous comparison submodule, which, in response to the regular monitoring instruction activating the sensor node, acquires a current real-time pollutant concentration reading, indexes the smooth fluctuation upper limit corresponding to the current time in the periodic environmental reference model, calculates a numerical deviation of the real-time pollutant concentration reading from the smooth fluctuation upper limit, and generates a real-time deviation value; an overrun degree evaluation submodule, which judges whether the real-time deviation value is positive, and if so, iteratively calculates a percentage overshoot of the real-time deviation value relative to the smooth fluctuation upper limit, compares the percentage overshoot with a preset fluctuation tolerance, screens an abnormal point with statistical significance, and generates an abnormal fluctuation signal; a sampling frequency triggering submodule, which, in response to the abnormal fluctuation signal, immediately locks a current monitoring time window, adjusts the sampling frequency to a preset maximum value, generates the potential deviation event including the current timestamp and the abnormal fluctuation signal, and assembles a high-frequency sampling instruction.

[0010] As a further scheme of the present application, the specific functions of the emergency collaborative control module are implemented as: The high-frequency data buffering sub-module receives the potential deviation event identifier, opens a cache area to store subsequent high-frequency pollutant concentration readings, establishes a sliding confirmation window with the current time as the starting point, counts the total number of all high-frequency pollutant concentration readings in the sliding confirmation window, and generates a window sample count value; The continuous verification sub-module traverses the high-frequency pollutant concentration readings in the sliding confirmation window, compares them one by one with the upper limit of the smooth fluctuation, accumulates the number of readings that exceed the upper limit of the smooth fluctuation, calculates the proportion of the readings in the window sample count value, and generates a confirmed deviation signal if the proportion exceeds the confirmation threshold; The device precise regulation sub-module responds to the confirmed deviation signal, retrieves the node-device association table to lock the environmental protection management device ID of the affected area, matches the preset management strategy table according to the pollution intensity level carried in the confirmed deviation signal, determines the target operating power and fan speed parameters, and generates a full-power operation environmental protection device regulation instruction.

[0011] As a further scheme of the present application, the specific process of generating the time point statistical feature vector in the statistical feature calculation sub-module includes: A standard time axis with a period of 24 hours is established, the obtained concentration data set after cleaning is mapped to a plurality of discrete time points on the standard time axis, and a time-aligned sample matrix is formed; For each discrete time point column vector in the sample matrix, a weighted average algorithm is applied to calculate the central tendency value of the time point, giving higher weight to recent data to improve the sensitivity to environmental changes, and obtaining the value of the daily mean reference curve at the corresponding time; For each discrete time point column vector in the sample matrix, the dispersion degree of all sample values relative to the central tendency value of the time point is calculated, the variance is calculated using an unbiased estimation formula, and the standard deviation is obtained by taking the square root of the variance, to obtain the value of the standard deviation at the corresponding time; The daily mean reference curve value and the standard deviation value calculated at the same time are combined, and the time attribute of the time point is marked through metadata, to generate the time point statistical feature vector.

[0012] As a further scheme of the present application, the specific process of constructing the low-power sleep instruction in the sleep strategy formulation sub-module includes: The value of the safety margin indicator and the percentage of the remaining battery capacity of the sensor node are obtained, and a sleep duration calculation model is established; According to the sleep duration calculation model, the sleep interval duration is nonlinearly increased with the decrease of the percentage of the remaining battery capacity within the risk range allowed by the safety margin indicator, to prolong the network survival cycle; calculating an absolute time at which the next wake-up for environment sampling is required, converting the absolute time into a countdown count value relative to the current system time; configuring a wireless communication module of the sensor node into a deep sleep mode, and writing the countdown count value into a hardware timer register of the node, generating the low-power hibernation instruction including the deep sleep mode configuration and the countdown count value.

[0013] As a further scheme of the present application, the specific process of generating the confirmed deviation signal in the persistent verification submodule includes: calculating a cumulative deviation intensity of the high-frequency pollutant concentration readings relative to the stationary fluctuation upper bound by using a cumulative risk integral formula, and performing a weighted evaluation combined with the number of over-limit readings, and generating the confirmed deviation signal if the calculation result exceeds a preset risk trigger line; The cumulative risk integral formula is specifically: ; wherein, represents a cumulative deviation intensity value on which the confirmed deviation signal is determined, represents a total number of samples in the sliding confirmation window, represents an i-th high-frequency pollutant concentration reading, represents a stationary fluctuation upper bound corresponding to the i-th sampling time, represents a time interval of high-frequency sampling, represents a preset persistent weight coefficient, represents a number of readings in the window that exceed the stationary fluctuation upper bound.

[0014] As a further scheme of the present application, the specific process of generating the full-power running environmental protection device control instruction in the device precise control submodule includes: analyzing the confirmed deviation signal to obtain a maximum value of the high-frequency pollutant concentration readings at the current time and an over-limit duration; According to the node-device association table, all associated environmental protection devices covering the potential deviation event occurrence area are identified, and the rated power parameters and the current running state of the multiple associated environmental protection devices are obtained; Based on the difference ratio between the maximum value of the high-frequency pollutant concentration readings and the stationary fluctuation upper bound, a pollution severity level is divided, and the pollution severity level is mapped to a load percentage of the device; If the load percentage reaches a full-load standard, the running mode is set to an emergency mode, the power level is set to 100% rated power, and the full-power running environmental protection device control instruction including the device start sequence and the full-speed running parameters is generated.

[0015] ​Compared with the prior art, the application has the advantages and positive effects that: In the application, by constructing a periodic environmental benchmark model and establishing an upper limit of smooth fluctuation, the environmental state is autonomously identified and the working mode is adjusted, the sleep state is switched when the data is below the threshold, the energy consumption is greatly reduced, the high-frequency sampling is triggered when the data breaks through the upper limit, the abnormality is accurately captured, the super-limit reading is confirmed twice, the device is automatically associated and activated to run at full power, the hierarchical response mechanism based on data driving discards manual intervention, realizes seamless connection from low-power inspection to emergency disposal, guarantees long-term endurance of the device and ensures real-time and efficient blocking of sudden pollution events. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 It is a functional module schematic diagram of the environmental protection equipment monitoring and regulation system of the application. Figure 2 It is a periodic environmental benchmark model construction flowchart of the application. Figure 3 It is a low-power inspection scheduling flowchart of the application. Figure 4 It is a sudden deviation detection flowchart of the application. Figure 5 It is an emergency cooperative regulation flowchart of the application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical scheme and advantages of the application clearer, the technical scheme realized based on software is described in detail below in combination with the system architecture diagram and the embodiment. It should be understood that the specific embodiments described herein are only used to explain the technical scheme of the application, and do not constitute a limitation on the protection scope.

[0018] In the description of the application, the system architecture relationship or data processing flow indicated by the terms "level", "module", "interface", "data flow", "client", "server" and the like are defined based on the corresponding architecture diagram or flowchart of the embodiment. This way of expression is only used to clearly explain the logical relationship of each element in the technical scheme, and is not limited to the physical deployment form. The "multiple" contains two or more technical units, including but not limited to multiple data nodes, processing threads, service instances or functional components, and other expandable elements, and the specific number is determined according to the actual business scenario.

[0019] Please refer to Figure 1 and Figure 2 , the application provides a technical scheme: an environmental protection equipment monitoring and regulation system based on a sensor network comprises: The quasi-model construction module collects the pollutant concentration data of the sensor node, calculates the mean value to form a daily mean value benchmark curve and a standard deviation, superimposes three times the standard deviation on the daily mean value benchmark curve to obtain a smooth fluctuation upper limit, combines the daily mean value benchmark curve and the smooth fluctuation upper limit to generate a periodic environmental benchmark model, and transmits the model to the low-power inspection scheduling module and the sudden deviation detection module; The periodic environmental benchmark model includes the daily mean value benchmark curve and the smooth fluctuation upper limit. The specific function of the environmental benchmark model construction module is as follows: The historical data cleaning submodule obtains the pollutant concentration data of the sensor node in a preset historical period, removes outliers in the pollutant concentration data caused by sensor failure, extracts valid time series samples and aligns them according to timestamps, and generates a cleaned concentration data set. The statistical feature calculation submodule calculates the arithmetic mean value of all samples for each time point based on the cleaned concentration data set to construct a daily mean value benchmark curve, calculates the standard deviation of the sample distribution at the same time point, quantifies the fluctuation range of environmental background noise according to the normal distribution principle, and generates a time point statistical feature vector. The specific process of generating the time point statistical feature vector in the statistical feature calculation submodule includes: A standard time axis with a period of twenty-four hours is established, and the obtained cleaned concentration data set is mapped to a plurality of discrete time points on the standard time axis to form a sample matrix after time alignment. For each discrete time point column vector in the sample matrix, the weighted average algorithm is applied to calculate the central tendency value of the time point, giving higher weight to recent data to improve the sensitivity to environmental changes, and obtaining the value of the daily mean value benchmark curve at the corresponding time; For each discrete time point column vector in the sample matrix, the dispersion degree of all sample values relative to the central tendency value of the time point is calculated, and the unbiased estimation formula is used to calculate the variance and take the square root to obtain the value of the standard deviation at the corresponding time. The daily mean value benchmark curve value and the standard deviation value calculated at the same time are combined, and the time attribute of the time is marked through metadata to generate a time point statistical feature vector. The dynamic boundary synthesis submodule fuses the daily mean value benchmark curve and the weighted standard deviation based on the time point statistical feature vector, introduces a historical extreme value fluctuation rate to adaptively correct the boundary, and uses a dynamic boundary synthesis formula: ; The smooth fluctuation upper limit of each time point is calculated, the daily mean value benchmark curve and the smooth fluctuation upper limit are combined, and a periodic environmental benchmark model is generated. Wherein, represents the smooth fluctuation upper limit at time t, This represents the value on the daily average baseline curve at time t. This represents the pre-defined coefficient of three standard deviations. This represents the standard deviation at time t. This represents the global maximum value of pollutant concentration data within a historical period. This represents the global minimum value of pollutant concentration data within a historical period. This represents the global average value of pollutant concentration data over a historical period.

[0020] The historical data cleaning submodule acquires pollutant concentration data from sensor nodes within a preset historical period. This preset historical period is set to 30 days. Taking a PM2.5 sensor node as an example, it acquires pollutant concentration data every 10 minutes over 30 days. The acquired pollutant concentration data is then inspected to remove outliers caused by sensor malfunctions. For example, the sensor's measurement range is 0-500. If a value of 999 is detected in the data or -10 If a reading is found to be outlier, it is identified as noise and removed. After removing outliers, valid time-series samples are extracted. If data for a 10-minute timestamp is missing, linear interpolation of the two preceding and following valid timestamps is used to complete the data. All valid samples are aligned according to their timestamps to generate a cleaned concentration dataset.

[0021] The statistical feature calculation submodule establishes a standard time axis with a 24-hour cycle. The discrete time points on this time axis are spaced 10 minutes apart, totaling 144 time points. The cleaned concentration dataset acquired over a 30-day period is mapped to the standard time axis according to the time points, forming a 30×144 time-aligned sample matrix. Each discrete time point column vector in the sample matrix (e.g., 30 data samples at 9:00 AM) is then processed.

[0022] Table 1: Subset of Concentration Data After Cleaning at 9:00 AM

[0023] Table 1 lists the post-wash concentration data at 9:00 AM for the past 5 days. A weighted average algorithm was used to calculate the central trend value at this time point. To improve sensitivity to recent environmental changes, more recent data were assigned higher weights. Weighting: A linear decaying weighting was used. The weight of the heaven ,in Total number of days (in this example) Taking the 5 days of data in Table 1 as an example () ), with weights respectively The daily average baseline curve at 9:00 AM. The calculation process is as follows: .

[0024] For the column vector of the 9:00 AM time point in the sample matrix, calculate the trend value of all sample values ​​(taking the 5 samples in Table 1 as an example) relative to the central trend value of the time point. The degree of dispersion of the variance is determined. The variance is calculated using an unbiased estimation formula. And take the square root to obtain the standard deviation. .

[0025] ; ; The daily average baseline curve values ​​calculated at the same time point. and standard deviation value Combine and tag the time attribute (9:00AM) of the time point by metadata to generate the statistical feature vector [Time:9:00,C_avg:37.2,Sigma_std:3.63].

[0026] The dynamic boundary synthesis submodule, based on the statistical feature vectors of all time points, fuses the daily average baseline curve and the weighted standard deviation to calculate the upper bound of the stationary fluctuation at each time point. The dynamic boundary synthesis formula is as follows: The parameters in the formula and their operational logic are explained below: For a moment The upper bound of the steady fluctuation; For a moment The daily average baseline curve value; The standard deviation coefficient is the preset value. For a moment Standard deviation; , , These represent the global maximum, global minimum, and global average values ​​of all cleaned concentration datasets within a preset historical period (30 days). The calculation logic of this formula is as follows: based on the daily average value... Based on this, a dynamically adjusted fluctuation boundary is added. This boundary is based on the standard deviation. Multiply by a coefficient This represents the statistical range of fluctuation. Simultaneously, we introduce... As a volatility factor, among which It is the ratio of the historical global fluctuation range to the global mean, used to quantify the volatility of the overall environment. When historical volatility is high, this factor increases, and the upper bound of stable fluctuations rises accordingly.

[0027] parameter Setting: represents the preset three standard deviation coefficient, whose value is set to 3. The setting of this value is based on the following experimental verification: 1000 different cities, different seasons of PM2.5 historical data set, a total of 1,000,000 data points. By normality test (such as Shapiro-Wilk test) on each data set, it is found that at the confidence level of 95%, about 92% of the data set sample distribution approximately obeys the normal distribution. In these approximately normally distributed data sets, the proportion of data points falling within the interval is statistically 99.73%. Based on this statistical result, set , adopt criterion.

[0028] Parameter assignment and example: bring in the statistical characteristic vector and global statistical value of 9:00 AM time point for calculation: (The foregoing calculation result) (The foregoing calculation result) (Has been set) : by traversing the 30-day all-washed concentration data set, the global maximum value . : by traversing the 30-day all-washed concentration data set, the global minimum value . : by calculating the arithmetic mean of the 30-day all-washed concentration data set, .

[0029] Calculate volatility factor: ; Calculate ; ; Take two decimal places, .

[0030] The advantage of the formula is that by introducing the historical extreme volatility This item, the standard deviation is dynamically weighted, so that the smooth volatility upper bound not only reflects the local volatility of the current time , but also adaptively integrates the global volatility characteristics of the entire historical period.

[0031] The result shows that at 9:00 AM, the smooth volatility upper bound of PM2.5 concentration is determined as This value is the daily average value reference curve ​The combination generates model data of the 9:00 AM time point in the periodic environment benchmark model. The above calculation is performed for all 144 time points, and finally a complete periodic environment benchmark model is generated.

[0032] Please refer to Figure 1 and Figure 3 The low-power inspection scheduling module calls the periodic environment benchmark model to obtain the upper bound of the smooth fluctuation, compares the upper bound of the smooth fluctuation with the device startup threshold, and if it is lower, establishes a low-power sleep instruction, and if it is higher, establishes a regular monitoring instruction, and synthesizes a sensor scheduling instruction. The sensor scheduling instruction includes a low-power sleep instruction and a regular monitoring instruction. The specific function implementation of the low-power inspection scheduling module is as follows: The safety margin analysis submodule analyzes the periodic environment benchmark model, extracts the upper bound of the smooth fluctuation sequence in the future one scheduling period, point-by-point calculates the difference between the upper bound of the smooth fluctuation sequence and the device startup threshold, selects the smallest positive difference as the safety margin index, and generates a periodic safety evaluation report. The sleep strategy formulation submodule calculates the sleep interval length in inverse proportion to the size of the safety margin index when the safety margin index is positive and greater than the preset safety buffer bandwidth, plans the time node of the next wake-up, and constructs a low-power sleep instruction. The specific process of constructing a low-power sleep instruction in the sleep strategy formulation submodule includes: Obtain the numerical value of the safety margin index and the battery remaining capacity percentage of the sensor node, and establish a sleep duration calculation model. According to the sleep duration calculation model, within the risk range allowed by the safety margin index, the sleep interval length is nonlinearly increased with the decrease of the battery remaining capacity percentage, so as to prolong the network survival cycle. Calculate the absolute time of the next time when the sensor node must be woken up for environment sampling, and convert the absolute time into a countdown count value relative to the current system time. Configure the wireless communication module of the sensor node to enter a deep sleep mode, and write the countdown count value into the hardware timer register of the node to generate a low-power sleep instruction including the deep sleep mode configuration and the countdown count value. The monitoring task distribution submodule sets a high-density regular sampling frequency when the safety margin index is less than or equal to the preset safety buffer bandwidth, configures the time slot parameters of data transmission, constructs a regular monitoring instruction, encapsulates the low-power sleep instruction or the regular monitoring instruction, and generates a sensor scheduling instruction.

[0033] Safety margin analysis submodule, analyze periodic environmental benchmark model, extract the upper bound sequence of smooth fluctuation in the future one scheduling period (for example, 1 hour). Assuming that the current time is 9:00 AM, extract the upper bound sequence of smooth fluctuation from 9:00 to 10:00 from the model, a total of 7 time points (interval 10 minutes): (unit: ). Obtain the device start threshold of the environmental protection equipment. The threshold is set according to the 24-hour average concentration limit value of PM2.5 secondary standard in "Environmental Air Quality Standard" (GB3095-2012) . . Calculate the difference between the upper bound sequence of smooth fluctuation and the device start threshold point by point : (unit: ). Screen the smallest positive difference as the safety margin index. The positive difference set is . The minimum value is . Generate periodic safety evaluation report, which contains safety margin index .

[0034] Hibernation strategy making submodule, compare the safety margin index with the preset safety buffer width. The safety buffer width is set as follows: the width is set according to the maximum change rate of environmental concentration under the data acquisition response delay of the sensor node and the maximum delay of network transmission . , . In the historical monitoring data, the maximum change rate of concentration is set to . . Set . The above maximum change rate of concentration refers to the maximum value of the absolute value of the concentration change rate between any two adjacent 10-minute sampling points in the preset historical period, that is , wherein is 10 minutes (600 seconds), and are the corresponding concentration values.

[0035] Comparison and judgment: , . Less than , does not meet the condition of "greater than the preset safety buffer width". (To complete the embodiment, the execution process when the condition is met is supplemented here): assuming another scene, the safety margin index . , meet the condition. Get the percentage of the remaining battery capacity of the sensor node, . Establish the sleep duration calculation model. The basic sleep duration of this model is 10 minutes, and is adjusted according to the safety margin and the battery capacity. The sleep interval duration is The calculation formula is: ; Wherein, is the calculated sleep interval duration of the sensor node; is the basic sleep duration; is the safety margin index; is the safety buffer width; is the percentage of the remaining battery capacity of the sensor node; the constant 0.5 is the battery capacity adjustment coefficient, which is used to set the influence degree of insufficient battery capacity on the sleep duration adjustment. . Calculate the absolute time when the sensor node must wake up for the next environment sampling. If the current time is 9:01:00 AM, the wake-up time is 9:01:00+960s=9:01:00+16 minutes=9:17:00 AM. Convert 960s to countdown count value relative to the current system time. Configure the wireless communication module (such as the NB-IoT module) of the sensor node to enter the deep sleep mode (PSM), and write the countdown count value 960 into the wake-up register of the hardware timer (such as the RTC) of the node. Generate the low-power sleep instruction [CMD:0x00, VALUE:960] including the deep sleep mode configuration and the countdown count value.

[0036] Monitor the task distribution sub-module, (return to the scenario) , . Less than , meet the condition of "safety margin index less than or equal to the preset safety buffer width". Set the high-density regular sampling frequency. The standard sampling frequency is 10 minutes, and the high-density sampling frequency is set to 1 minute (60 seconds). Configure the time slot parameters of data transmission. In the TDMA network, allocate the 100ms time slot starting at the 5th second of each minute for the node to perform data transmission. Build a regular monitoring instruction, and the instruction content includes the sampling frequency 60s and the time slot parameter [Freq:60, Slot:5]. Encapsulate the regular monitoring instruction to generate the sensor scheduling instruction [CMD:0x01, Freq:60, Slot:5].

[0037] Please refer to Figure 1 and Figure 4, a sudden deviation detection module, monitoring real-time pollutant concentration readings, if the real-time pollutant concentration readings exceed the upper limit of the periodic environmental benchmark model, a potential deviation event is determined, and a high-frequency sampling instruction is generated and transmitted to the emergency coordination control module; The high-frequency sampling instruction includes a potential deviation event identifier and a high-frequency pollutant concentration reading collection; The specific function of the sudden deviation detection module is: A real-time synchronization comparison submodule, in response to a regular monitoring instruction, activates the sensor node, obtains the current real-time pollutant concentration reading, indexes the upper limit of the periodic environmental benchmark model corresponding to the current time, calculates the numerical deviation of the real-time pollutant concentration reading from the upper limit of the periodic environmental benchmark model, and generates a real-time deviation value; An over-limit degree evaluation submodule, determines whether the real-time deviation value is positive, if it is positive, iteratively calculates the percentage overshoot of the real-time deviation value relative to the upper limit of the periodic environmental benchmark model, compares the percentage overshoot with the preset fluctuation tolerance, filters statistically significant abnormal points, and generates an abnormal fluctuation signal; A sampling frequency triggering submodule, in response to the abnormal fluctuation signal, immediately locks the current monitoring time window, adjusts the sampling frequency to a preset maximum value, generates a potential deviation event including the current timestamp and the abnormal fluctuation signal, and assembles the high-frequency sampling instruction.

[0038] A real-time synchronization comparison submodule, in response to the sensor scheduling instruction [CMD: 0x01, Freq: 60, Slot: 5] generated by the low-power inspection scheduling module, the sensor node is activated and sampled at a frequency of 1 minute. At 9:15:00 AM, the current real-time pollutant concentration reading is obtained . Indexes the upper limit of the periodic environmental benchmark model corresponding to the current time 9:15:00 AM (assuming it is in the 9:10-9:20 interval). The query from the model obtains . Calculate the numerical deviation of the real-time pollutant concentration reading from the upper limit of the periodic environmental benchmark model . . Generate real-time deviation value .

[0039] An over-limit degree evaluation submodule, determines whether the real-time deviation value is positive. 5.5>0, positive. Iteratively calculate the percentage overshoot of the real-time deviation value relative to the upper limit of the periodic environmental benchmark model . . Compare the percentage overshoot 7.097% with the preset fluctuation tolerance . The setting of the fluctuation tolerance : this tolerance According to the historical data, the instantaneous, small, and non-continuous fluctuations caused by non-pollution sources (such as gusts and passing vehicles) are statistically set. Statistics show that 98% of such instantaneous fluctuations cause overshoots within 5%. To balance sensitivity and false alarm rate, the overshoot is set to 5.0%. Comparison: 7.097%>5.0%. Filter the statistically significant abnormal points to generate an abnormal fluctuation signal [Signal:0x01].

[0040] The sampling frequency triggering submodule responds to the abnormal fluctuation signal [Signal:0x01]. The current monitoring time window 9:15:00 AM is immediately locked. The sampling frequency is adjusted to the preset maximum value. The frequency of the regular monitoring instruction is 60 seconds, and the preset maximum value is set to 10 seconds. A potential deviation event [EventID:E1234,Timestamp:9:15:00,Signal:0x01] is generated, including the current timestamp 9:15:00 AM and the abnormal fluctuation signal. The high-frequency sampling instruction is assembled, including the event ID and the maximum frequency 10s, [CMD:0x02,EventID:E1234,Freq:10].

[0041] Please refer to Figure 1 and Figure 5 , the emergency cooperative regulation module monitors the high-frequency pollutant concentration readings based on the potential deviation event. If the number of high-frequency pollutant concentration readings exceeding the upper bound of the stationary fluctuation meets the confirmation threshold, the node-device association table is searched, and a full-power operation of environmental protection equipment control instruction is generated; The full-power operation of environmental protection equipment control instruction includes device identification, operation mode, power level, etc. The specific function implementation of the emergency cooperative regulation module is as follows: The high-frequency data buffer submodule receives the potential deviation event identifier, opens a high-speed cache area to store subsequent high-frequency pollutant concentration readings, establishes a sliding confirmation window with the current time as the starting point, counts the total number of all high-frequency pollutant concentration readings in the sliding confirmation window, and generates a window sampling count value; The persistence verification submodule traverses the high-frequency pollutant concentration readings in the sliding confirmation window, compares them one by one with the upper bound of the stationary fluctuation, accumulates the number of readings exceeding the upper bound of the stationary fluctuation, calculates the proportion of the readings exceeding the upper bound of the stationary fluctuation in the window sampling count value, and generates a confirmed deviation signal if the proportion exceeds the confirmation threshold; The specific process of generating a confirmed deviation signal in the persistence verification submodule includes: The cumulative risk integral formula is used to calculate the cumulative deviation intensity of the high-frequency pollutant concentration readings relative to the upper bound of the stationary fluctuation, and the number of over-limit readings is used for weighted evaluation. If the calculation result exceeds the preset risk trigger line, a confirmed deviation signal is generated; ​The cumulative risk score formula is specifically: ; wherein, represents the cumulative deviation intensity value on which the confirmed deviation signal judgment is based, N represents the total number of samples in the sliding confirmation window, represents the i-th high-frequency pollutant concentration reading, represents the i-th sampling time corresponding to the upper limit of the smooth fluctuation, represents the time interval of high-frequency sampling, represents the preset persistence weight coefficient, represents the number of readings in the window that exceed the upper limit of the smooth fluctuation; The device precise regulation submodule responds to the confirmed deviation signal, retrieves the node-device association table to lock the environmental protection management device ID of the affected area, matches the preset management strategy table according to the pollution intensity level carried in the confirmed deviation signal, determines the target operating power and fan speed parameters, and generates a full-power operation environmental protection device regulation instruction; The specific process of generating a full-power operation environmental protection device regulation instruction in the device precise regulation submodule includes: Analyzing the confirmed deviation signal to obtain the maximum value of the current high-frequency pollutant concentration reading and the over-limit duration; According to the node-device association table, identify all associated environmental protection devices covering the potential deviation event occurrence area, and obtain the rated power parameters and current operating state of the multiple associated environmental protection devices; Based on the difference ratio of the maximum value of the high-frequency pollutant concentration reading and the upper limit of the smooth fluctuation, the pollution severity level is divided, and the pollution severity level is mapped to the load percentage of the device; If the load percentage reaches the full load standard, set the operating mode to emergency mode, set the power level to 100% rated power, and generate a full-power operation environmental protection device regulation instruction including the device start sequence and full-speed operation parameters.

[0042] The high-frequency data buffering submodule receives the potential deviation event identifier EventID:E1234 from the sudden deviation detection module. A cache area is opened with a capacity of 6 samples for storing subsequent high-frequency pollutant concentration readings. A sliding confirmation window is established with the current time 9:15:00 AM as the starting point. The window length is set to 1 minute. The high-frequency sampling frequency is 10 seconds, and the total number of samples in the window is 6. During 9:15:10 to 9:16:00, 6 high-frequency pollutant concentration readings are collected and stored, and the corresponding upper limit of the smooth fluctuation at each time is also obtained.

[0043] Table 2: High-frequency data table of sliding confirmation window

[0044] As shown in Table 2, the total number N = 6 of all high-frequency pollutant concentration readings in the sliding confirmation window is counted, and a window sample count value 6 is generated.

[0045] The persistence verification submodule traverses the high-frequency pollutant concentration readings in the sliding confirmation window (see Table 2) and compares them one by one with the upper bound of smooth fluctuation. (overrun) (overrun) (overrun) (overrun) (overrun) (overrun) cumulative number of readings exceeding the upper bound of smooth fluctuation The proportion of the cumulative number of readings exceeding the upper bound of smooth fluctuation in the window sample count value is calculated The confirmation threshold is obtained The threshold is set to 50%, that is, at least half of the data in the window is considered to have persistence. , the proportion condition is met. The cumulative deviation intensity of the high-frequency pollutant concentration readings with respect to the upper bound of smooth fluctuation is calculated using the cumulative risk integral formula: ; wherein, is the cumulative deviation intensity value; is the total number of samples in the window; is the serial number of the sample; is the high-frequency pollutant concentration reading; is the sampling time corresponding to the upper bound of smooth fluctuation; is the time interval of high-frequency sampling; is the persistence weight coefficient, used to amplify the risk level of high-persistence events; is the number of readings exceeding the window. The operation logic of this formula is: the sum of each overrun amount in the window is summed up, and multiplied by the time interval to calculate a cumulative "concentration-time" integral representing the total flux of pollution. At the same time, an exponential weighting term is introduced, which reflects the persistence of the overrun. When the persistence is stronger (the is closer to 1), the exponential weight is larger, and the value is amplified to a higher degree.

[0046] Setting of parameters and : (Already set). Persistence weight coefficient : Fitted by experimental data, when , weight when , weight when . This setting can effectively amplify the risk level of high persistence events. Set .

[0047] Parameter assignment and example (according to Table 2 data): ; Calculate the sum term :

[0048] ; ; ; ; ; ; Calculate the exponential weighted term: ; Calculate : ; (unit: ) Compare the result with the preset risk trigger line . The setting of the risk trigger line : This value represents the minimum cumulative pollution amount that needs to start the emergency control. According to the control experience, set the event of "over-limit for 1 minute (60 seconds)" as the trigger benchmark. (Note: This benchmark does not consider exponential weighting, which is a conservative benchmark) (use the benchmark considering weighting) (unit: ). To improve response sensitivity, take 90% of the benchmark value, . Set . Comparison: . The calculation result exceeds the preset risk trigger line, generating a confirmation deviation signal [Signal: 0x02, EventID: E1234, Risk: 1573.9].

[0049] The benefit of the formula is that it calculates the over-limit amplitude by , accumulates the time by , and accumulates the risk by A persistent nonlinear penalty was introduced, making It can quantify the overall risk of sudden pollution events more accurately than simple threshold comparisons.

[0050] The results indicate that within the 1-minute window of 9:15-9:16, the cumulative deviation intensity of 1573.9 exceeded the risk trigger line of 1450, confirming a statistically significant and actually harmful persistent deviation event.

[0051] The equipment precision control submodule responds to the confirmation deviation signal [Signal:0x02,EventID:E1234,Risk:1573.9]. It retrieves the node-equipment association table.

[0052] Table 3: Node-Device Association Table

[0053] As shown in Table 3, assuming event E1234 originates from sensor node SN-PM25-008, located in plant area B, the search results are SPRAY-SYS-02 and FAN-SYS-03. The deviation signal is analyzed to obtain the maximum value of the high-frequency pollutant concentration reading at the current moment. (See Table 2, i=3), duration of exceeding limit Based on the node-equipment association table, identify all associated environmental protection equipment SPRAY-SYS-02 and FAN-SYS-03 covering the area where potential deviation events may occur. Obtain their rated power (50kW for the spray tower and 15kW for the auxiliary fan) and current operating status (standby). Classify the pollution severity level based on the ratio of the difference between the maximum value of the high-frequency pollutant concentration reading and the upper limit of the steady fluctuation. Pollution severity level classification: Level 1: (Starting at 50% power) Level 2: (Starting at 100% power) Level 3: (Start at 100% power and alarm) 16.97% belongs to Level 2. The load percentage has reached the full load standard (100%). Set the operating mode of SPRAY-SYS-02 and FAN-SYS-03 to emergency mode Mode:0x03, and set the power level to 100% rated power Power:100. Generate full-power operation environmental protection equipment control instructions including equipment start-up sequence and full-speed operation parameters, instruction 1: [Target:SPRAY-SYS-02,Mode:0x03,Power:100], instruction 2: [Target:FAN-SYS-03,Mode:0x03,Power:100].

[0054] The node-device association table refers to a database stored in the system and used to establish the geographical and functional correspondence between the sensor node and the environmental protection device to be regulated.

[0055] The above embodiments demonstrate the preferred embodiments of the present application, and any equivalent adjustment of the technical solutions based on software engineering methods is within the protection scope, including but not limited to: implementing algorithm logic in different programming languages, service reconstruction of functional modules, adjustment of data interaction protocol, optimization of resource scheduling strategy and other technical improvements. Any implementation derived by reasonable modification of the data processing flow, service calling link or system architecture level without deviating from the technical core of the present application should be considered within the protection scope defined in the claims of the present application.

Claims

1. A kind of environmental protection equipment monitoring and control system based on sensor network, it is characterized in that, The system comprises: An environmental benchmark model construction module, which collects pollutant concentration data of a sensor node, calculates a mean value to form a daily mean benchmark curve and a standard deviation, superimposes three times the standard deviation on the daily mean benchmark curve to obtain a smooth fluctuation upper limit, combines the daily mean benchmark curve and the smooth fluctuation upper limit to generate a periodic environmental benchmark model, and transmits the model to a low-power inspection scheduling module and a sudden deviation detection module; The low-power inspection scheduling module calls the periodic environmental benchmark model and the smooth fluctuation upper limit, compares the smooth fluctuation upper limit with a device start threshold, establishes a low-power sleep instruction if the smooth fluctuation upper limit is lower than the device start threshold, establishes a regular monitoring instruction if the smooth fluctuation upper limit is higher than the device start threshold, and synthesizes a sensor scheduling instruction; The sudden deviation detection module monitors real-time pollutant concentration readings, determines a potential deviation event if the real-time pollutant concentration readings exceed the smooth fluctuation upper limit of the periodic environmental benchmark model, generates a high-frequency sampling instruction, and transmits the instruction to an emergency collaborative control module; The emergency collaborative control module monitors high-frequency pollutant concentration readings based on the potential deviation event, checks a node-device association table if the number of high-frequency pollutant concentration readings exceeding the smooth fluctuation upper limit meets a confirmation threshold, and generates a full-power operation of environmental protection equipment control instruction. 2.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 1, wherein, The periodic environmental benchmark model comprises a daily mean benchmark curve and a smooth fluctuation upper limit, the sensor scheduling instruction comprises a low-power sleep instruction and a regular monitoring instruction, the high-frequency sampling instruction comprises a potential deviation event identifier and high-frequency pollutant concentration reading collection, and the full-power operation of environmental protection equipment control instruction comprises a device identifier, an operation mode, and a power level. 3.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 2, wherein, The environmental benchmark model construction module is specifically implemented as follows: A historical data cleaning submodule acquires pollutant concentration data of a sensor node in a preset historical period, removes outliers in the pollutant concentration data caused by sensor failure, extracts valid time series samples and aligns them according to timestamps, and generates a cleaned concentration data set; A statistical feature calculation submodule calculates, based on the cleaned concentration data set, an arithmetic mean value of all samples for each time point to construct a daily mean benchmark curve, calculates a standard deviation of sample distribution at the same time point, quantifies a fluctuation range of environmental background noise according to a normal distribution principle, and generates a time point statistical feature vector; A dynamic boundary synthesis submodule calculates, based on the time point statistical feature vector, a smooth fluctuation upper limit for each time point, combines the daily mean benchmark curve and the smooth fluctuation upper limit, and generates a periodic environmental benchmark model; ； The low-power inspection scheduling module is specifically implemented as follows: wherein, represents the stationary fluctuation upper bound at time t, represents the value on the daily mean reference curve at time t, represents a preset three standard deviation coefficient, represents the standard deviation at time t, represents the global maximum value of the pollutant concentration data in the historical period, represents the global minimum value of the pollutant concentration data in the historical period, represents the global average value of the pollutant concentration data in the historical period. 4.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 3, wherein, A safety margin analysis submodule analyzes the periodic environmental benchmark model, extracts a smooth fluctuation upper limit sequence in a future scheduling period, calculates a difference between the smooth fluctuation upper limit sequence and a device start threshold point by point, screens a smallest positive difference as a safety margin indicator, and generates a periodic safety evaluation report; and A safety margin analysis submodule analyzes the periodic environmental benchmark model, extracts a smooth fluctuation upper limit sequence in a future scheduling period, calculates a difference between the smooth fluctuation upper limit sequence and a device start threshold point by point, screens a smallest positive difference as a safety margin indicator, and generates a periodic safety evaluation report. The hibernation strategy sub-module calculates a hibernation interval length in inverse proportion to the size of the safety margin index when the safety margin index is positive and greater than a preset safety buffer bandwidth, plans a time node for next wake-up, and constructs a low-power hibernation instruction; The monitoring task distribution sub-module sets a high-density regular sampling frequency, configures time slot parameters for data transmission, constructs a regular monitoring instruction, encapsulates the low-power hibernation instruction or the regular monitoring instruction, and generates a sensor scheduling instruction when the safety margin index is less than or equal to the preset safety buffer bandwidth.

5. The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 4, characterized in that, The specific function of the burst deviation detection module is implemented as: The real-time synchronous comparison sub-module activates the sensor node in response to the regular monitoring instruction, acquires a current real-time pollutant concentration reading, indexes the upper bound of the smooth fluctuation in the periodic environmental reference model corresponding to the current time, calculates a numerical deviation of the real-time pollutant concentration reading from the upper bound of the smooth fluctuation, and generates a real-time deviation value; The overrun degree evaluation sub-module judges whether the real-time deviation value is positive, iteratively calculates a percentage overshoot of the real-time deviation value relative to the upper bound of the smooth fluctuation if the real-time deviation value is positive, compares the percentage overshoot with a preset fluctuation tolerance, screens abnormal points with statistical significance, and generates an abnormal fluctuation signal; The sampling frequency triggering sub-module immediately locks a current monitoring time window, adjusts a sampling frequency to a preset maximum value, generates the potential deviation event including a current timestamp and the abnormal fluctuation signal, and assembles a high-frequency sampling instruction in response to the abnormal fluctuation signal. 6.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 2, wherein, The specific function of the emergency cooperative regulation module is implemented as: The high-frequency data buffering sub-module receives a potential deviation event identifier, opens a high-speed cache area to store subsequent high-frequency pollutant concentration readings, establishes a sliding confirmation window with the current time as a starting point, counts a total number of all high-frequency pollutant concentration readings in the sliding confirmation window, and generates a window sampling count value; The continuous verification sub-module traverses the high-frequency pollutant concentration readings in the sliding confirmation window, compares them one by one with the upper bound of the smooth fluctuation, accumulates the number of readings exceeding the upper bound of the smooth fluctuation, calculates the proportion of the readings in the window sampling count value, and generates a confirmed deviation signal if the proportion exceeds a confirmation threshold; The device precise regulation sub-module retrieves the node-device association table to lock an environmental protection management device ID of an affected area in response to the confirmed deviation signal, matches a preset management strategy table according to a pollution intensity level carried in the confirmed deviation signal, determines target operating power and fan speed parameters, and generates a full-power operation environmental protection device regulation instruction. 7.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 3, wherein, The specific process of generating a time point statistical feature vector in the statistical feature calculation sub-module includes: A standard time axis with a period of 24 hours is established, the obtained concentration data set after cleaning is mapped to a plurality of discrete time points on the standard time axis, and a time-aligned sample matrix is formed. For each discrete time point column vector in the sample matrix, a weighted average algorithm is applied to calculate the central tendency value of the time point, giving higher weight to recent data to improve the sensitivity to environmental changes, and obtaining the value of the daily average reference curve at the corresponding time point; For each discrete time point column vector in the sample matrix, the dispersion degree of all sample values relative to the central tendency value of the time point is calculated, and the variance is calculated using an unbiased estimation formula and then taking the square root to obtain the value of the standard deviation at the corresponding time point; The daily average reference curve value and the standard deviation value calculated at the same time are combined, and the time attribute of the time point is marked by metadata to generate the time point statistical feature vector. 8.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 4, wherein, The specific process of constructing a low-power sleep instruction in the sleep strategy formulation submodule includes: Obtain the value of the safety margin index and the percentage of the remaining battery capacity of the sensor node, and establish a sleep duration calculation model; According to the sleep duration calculation model, within the risk range allowed by the safety margin index, the sleep interval duration is nonlinearly increased with the decrease of the percentage of the remaining battery capacity to prolong the network survival cycle; Calculate the absolute time of the next time when the sensor node must be woken up for environmental sampling, and convert the absolute time into a countdown count value relative to the current system time; Configure the wireless communication module of the sensor node to enter a deep sleep mode, and write the countdown count value into the hardware timer register of the node to generate the low-power sleep instruction including the deep sleep mode configuration and the countdown count value. 9.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 6, wherein, The specific process of generating a confirmation deviation signal in the persistent verification submodule includes: Calculate the cumulative deviation intensity of the high-frequency pollutant concentration reading relative to the smooth fluctuation upper limit using the cumulative risk integral formula, and perform weighted evaluation combined with the number of over-limit readings. If the calculation result exceeds the preset risk trigger line, the confirmation deviation signal is generated; The cumulative risk integral formula is specifically: ； wherein, represents the cumulative deviation intensity value on which the deviation signal determination is based, represents the total number of samples within the sliding confirmation window, represents the i-th high-frequency pollutant concentration reading, represents the i-th sampling time instant corresponding stationary fluctuation upper bound, represents the time interval of high-frequency sampling, represents the preset persistence weight coefficient, represents the number of readings within the window that exceed the stationary fluctuation upper bound. 10.The environmental protection equipment monitoring and regulation system based on a sensor network according to claim 6, wherein, The specific process of generating a full-power running environmental protection equipment control instruction in the device precise control submodule includes: Parse the confirmation deviation signal to obtain the maximum value of the high-frequency pollutant concentration reading at the current time and the over-limit duration; According to the node-device association table, identify all associated environmental protection equipment covering the potential deviation event occurrence area, and obtain the rated power parameters and current running state of multiple associated environmental protection equipment; Based on the difference ratio between the maximum value of the high-frequency pollutant concentration reading and the smooth fluctuation upper limit, the pollution severity level is divided, and the pollution severity level is mapped to the load percentage of the device; If the load percentage reaches the full load standard, the running mode is set to emergency mode, the power level is set to 100% rated power, and the full-power running environmental protection equipment control instruction including the device start sequence and full-speed running parameters is generated.

Images