Rural sewage intelligent monitoring and aeration regulation and control system based on Internet of Things

Abstract

The invention relates to the technical field of rural sewage treatment, and discloses a rural sewage intelligent monitoring and aeration regulation and control system based on Internet of Things. The system comprises a pollution parameter sensing layer, a pollution index analysis layer, an aeration path generation layer, a process parameter matching layer, a dynamic monitoring focusing layer, a cross-modal mapping layer, an equipment scheduling decision-making layer and an aeration parameter generation layer. Wherein the pollution parameter sensing layer collects operation parameters in real time by means of a multi-source sensing terminal; the pollution index analysis layer divides pollution parameter intervals and calculates a distribution coefficient; the aeration path generation layer generates an aeration intensity grading path in sequence; the process parameter matching layer outputs a process deviation parameter set; dynamically monitoring a focusing layer to adjust the density of monitoring points; the cross-modal mapping layer establishes space-time correlation mapping of water quality parameters and aeration intensity; the equipment scheduling decision-making layer sorts the calling sequence of the aeration equipment; and the aeration parameter generation layer converts the parameters into executable regulation and control parameters. The system realizes intelligent monitoring of rural sewage and accurate aeration regulation and control.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rural sewage treatment, in particular to an intelligent monitoring and aeration control system for rural sewage based on the Internet of Things. BACKGROUND

[0002] Rural sewage treatment is an important part of improving the rural living environment, but its treatment process faces many challenges. At present, rural sewage treatment facilities are mostly distributed in remote areas with complex geographical environment and lack of professional operation and maintenance personnel, resulting in unstable sewage treatment effect. Traditional sewage treatment monitoring methods mostly rely on manual inspection, which not only consumes time and effort, but also has the problems of low monitoring frequency and data lag, making it difficult to real-time grasp the sewage quality change and treatment facility operation state. In terms of aeration control, the existing system mostly uses fixed parameters for operation, which cannot dynamically adjust the aeration intensity according to the actual concentration of pollutants in the sewage. When the concentration of sewage pollutants suddenly increases, the fixed aeration intensity is difficult to meet the treatment demand, which easily leads to substandard effluent quality; while when the concentration of pollutants is low, continuous high-intensity aeration will cause energy waste and increase operation cost. The monitoring point layout of rural sewage treatment facilities is usually fixed, and it is difficult to flexibly adjust the monitoring density according to the water quality change. In areas with large water quality fluctuations, fixed monitoring points may not be able to timely capture water quality abnormalities, affecting the accurate regulation and control of the sewage treatment process; while in areas with stable water quality, too many monitoring points will cause resource idling. The existence of these problems makes it difficult to improve the efficiency of rural sewage treatment, which restricts the improvement of rural ecological environment. SUMMARY

[0003] The present application aims to provide an intelligent monitoring and aeration control system for rural sewage based on the Internet of Things to solve the problems raised in the background.

[0004] To achieve the above-mentioned purpose, the present application provides an intelligent monitoring and aeration control system for rural sewage based on the Internet of Things, which comprises: A pollution parameter perception layer for real-time collection of operation parameters of sewage treatment facilities through multi-source sensing terminals; A pollution index analysis layer for obtaining operation parameters of sewage treatment facilities, dividing pollution parameter intervals according to operation parameters and calculating pollution index distribution coefficients; An aeration path generation layer for generating aeration intensity grading paths according to pollution index distribution coefficients; A process parameter matching layer for comparing current process parameters with preset process parameter thresholds and outputting a process deviation parameter set; A dynamic monitoring focusing layer for dynamically adjusting the monitoring point density of multi-source sensing terminals according to the process deviation parameter set; a cross-modal mapping layer, which synchronously receives the monitoring point distribution data of the dynamic monitoring focusing layer and the aeration intensity hierarchical path of the aeration path generating layer, and establishes a spatiotemporal correlation mapping of the water quality parameter and the aeration intensity; a device scheduling decision layer, which sorts the aeration device calling sequence according to the spatiotemporal correlation mapping result; an aeration parameter generating layer, which converts the aeration device calling sequence into executable aeration control parameters.

[0005] Preferably, the system further comprises: an aeration effect backtracking layer, which is connected to the aeration parameter generating layer, compares the water quality parameter variation before and after the execution of the aeration control parameters, and dynamically corrects the pollution parameter interval division logic according to the water quality parameter variation.

[0006] Preferably, the pollution index analysis layer comprises: obtaining the maximum pollution load range of the sewage treatment facility, and equally dividing the pollution load range into a plurality of parameter intervals; extracting the pollution type and concentration threshold in the operation parameter, and querying the corresponding pollution characteristic curve in the pollution characteristic table; obtaining the pollution index allocation coefficient of each parameter interval through a pollution index allocation coefficient calculation model; calculating the number of monitoring points corresponding to each parameter interval based on the pollution index allocation coefficient; equally arranging a plurality of key monitoring points in the corresponding parameter interval.

[0007] Preferably, the aeration path generating layer comprises: marking the key monitoring point with the highest pollution index allocation coefficient as a severe pollution monitoring point; marking the key monitoring point with the lowest pollution index allocation coefficient as a mild pollution monitoring point; obtaining the pollution value of the initial monitoring point corresponding to the pollution characteristic curve; judging whether the difference between the pollution value of the severe pollution monitoring point and the pollution value of the initial monitoring point is greater than the difference between the pollution value of the mild pollution monitoring point and the pollution value of the initial monitoring point; if yes, generating a first aeration path from the initial monitoring point to the severe pollution monitoring point, a second aeration path from the severe pollution monitoring point back to the initial monitoring point, and a third aeration path from the initial monitoring point to the mild pollution monitoring point; if no, generating a first aeration path from the initial monitoring point to the mild pollution monitoring point, a second aeration path from the mild pollution monitoring point back to the initial monitoring point, and a third aeration path from the initial monitoring point to the severe pollution monitoring point; integrating the first aeration path, the second aeration path and the third aeration path to form the aeration intensity hierarchical path.

[0008] Preferably, the process parameter matching layer comprises: Statistically match the absolute deviation of the current process parameters with the preset process parameter threshold value; Calculate the weighted sum of the dissolved oxygen deviation, biochemical oxygen demand deviation, and suspended matter deviation; When the weighted sum exceeds the process tolerance threshold value, mark the corresponding parameter interval as a process deviation parameter set.

[0009] Preferably, the dynamic monitoring focusing layer comprises: Identify the key monitoring points covered by the process deviation parameter set, and upgrade the area where the key monitoring points are located to a dense monitoring area; The area covered by the non-process deviation parameter set is downgraded to a sparse monitoring area, the dense monitoring area adopts a high-frequency sampling mode, and the sparse monitoring area adopts a low-frequency sampling mode.

[0010] Preferably, the cross-modal mapping layer comprises: Establish timestamp alignment between the water quality parameter change curve of the dense monitoring area and the aeration intensity grading path; Label the water quality parameter fluctuation characteristics corresponding to each node in the aeration intensity grading path, and generate an associated mapping table of water quality parameter fluctuation characteristics and aeration intensity levels.

[0011] Preferably, the device scheduling decision layer comprises: Extract the nodes in the associated mapping table where the aeration intensity level meets the standard, sort the aeration device call priority according to the node meeting duration, and combine the continuous meeting nodes into an aeration device efficient operation interval.

[0012] Preferably, the aeration parameter generation layer comprises: Convert the aeration device efficient operation interval into aeration fan speed ladder instructions, generate a blower start-stop sequence according to the aeration device call priority, and combine the aeration fan speed ladder instructions and the blower start-stop sequence to form executable aeration control parameters.

[0013] Preferably, the aeration effect backtracking layer comprises: Record the baseline water quality parameters of the key monitoring points before executing the executable aeration control parameters, collect the real-time water quality parameters of the same key monitoring points after execution, and calculate the absolute change amount of the real-time water quality parameters relative to the baseline water quality parameters; When the absolute change amount does not reach the expected improvement threshold value, increase the number of monitoring points in the corresponding parameter interval; When the absolute change amount continuously exceeds the expected improvement threshold value, reduce the number of monitoring points in the corresponding parameter interval; Update the pollution characteristic curve weight coefficient in the pollution characteristic reference table.

[0014] Compared with the prior art, the present application has the following advantages: The multi-source sensing terminal of the pollution parameter perception layer can collect the operation parameters of the sewage treatment facility in real time, so that the management personnel can know the changes of various indexes in the sewage treatment process in a timely manner, and the limitations of traditional manual monitoring are broken. The pollution index analysis layer divides the pollution parameter interval according to the operation parameters and calculates the distribution coefficient, which provides a scientific basis for subsequent aeration regulation, and makes the evaluation of the degree of sewage pollution more accurate. The aeration path generation layer generates an aeration intensity grading path according to the pollution index distribution coefficient, changes the previous fixed aeration intensity operation mode, can adjust the aeration strategy according to the actual situation of the pollutants in the sewage, and makes the aeration process more in line with the actual needs of sewage treatment. The process parameter matching layer compares the current process parameters with the preset threshold value, and outputs a process deviation parameter set, which is helpful to find problems existing in the sewage treatment process in a timely manner and provides a direction for process adjustment. The dynamic monitoring focusing layer dynamically adjusts the monitoring point density according to the process deviation parameter set, increases the monitoring points in the area with large water quality fluctuations, and reduces the monitoring points in the area with stable water quality, so as to realize the reasonable allocation of monitoring resources, ensure the timely capture of water quality abnormalities, and avoid waste of resources. The cross-modal mapping layer establishes the spatio-temporal correlation mapping between water quality parameters and aeration intensity, combines the monitoring point distribution data with the aeration intensity grading path, makes the relationship between the two more clear, and provides a more comprehensive reference for equipment scheduling. The equipment scheduling decision layer sorts the aeration equipment calling sequence according to the spatio-temporal correlation mapping result, so that the operation of the aeration equipment is more orderly, and the utilization efficiency of the equipment is improved. The aeration parameter generation layer converts the equipment calling sequence into executable regulation parameters, ensures the effective execution of the aeration regulation instruction, forms a closed loop for the whole aeration process, and improves the stability and reliability of the sewage treatment. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 The timing diagram of the rural sewage intelligent monitoring and aeration regulation system based on the Internet of Things is described. Figure 2 The flowchart of pollution index analysis is described. Figure 3 The flowchart of aeration path generation is described. Figure 4 The flowchart of process parameter matching is described. Figure 5 The flowchart of cross-modal mapping is described. DETAILED DESCRIPTION

[0016] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0017] With reference to Figure 1 The present application provides an intelligent monitoring and aeration control system for rural sewage based on the Internet of Things. The system comprises: through the cooperative action of a pollution parameter sensing layer, a pollution index analysis layer, an aeration path generation layer, a process parameter matching layer, a dynamic monitoring focusing layer, a cross-modal mapping layer, a device scheduling decision layer and an aeration parameter generation layer, the intelligent operation of a sewage treatment facility is realized.

[0018] The pollution parameter sensing layer collects the operation parameters of the sewage treatment facility in real time through a multi-source sensing terminal, including the key indicators such as dissolved oxygen concentration, biochemical oxygen demand and suspended solids concentration. The pollution index analysis layer divides the pollution parameter interval according to the operation parameters and calculates the pollution index distribution coefficient, which provides a quantitative basis for subsequent aeration control. The aeration path generation layer generates an aeration intensity grading path based on the pollution index distribution coefficient, which guides the operation strategy of the aeration equipment. The process parameter matching layer compares the current process parameters with the preset threshold value, and outputs a process deviation parameter set. The dynamic monitoring focusing layer adjusts the monitoring point density accordingly, and optimizes the data collection efficiency. The cross-modal mapping layer establishes the spatio-temporal correlation mapping of water quality parameters and aeration intensity. The device scheduling decision layer sorts the aeration equipment calling sequence according to the mapping result, and finally outputs the executable aeration control parameters by the aeration parameter generation layer, completing the closed-loop control.

[0019] Embodiment 1: With reference to Figure 2 Before the operation of the pollution index analysis layer, the multi-source sensing terminal has completed the parameter collection of the whole area of the sewage treatment facility. The operation parameters include dissolved oxygen concentration, biochemical oxygen demand (BOD), chemical oxygen demand (COD), suspended solids (SS) concentration, ammonia nitrogen content and acid-base value, etc. All data are encrypted and packaged through the Internet of Things transmission protocol, and transmitted to the central processing unit of the analysis layer. The maximum pollution load range preset by the system is set according to the design processing capacity of the sewage treatment facility. This range is divided into a fixed number or a dynamic number of parameter intervals. In the dynamic division mode, the number of intervals is automatically adjusted according to the real-time pollution load fluctuation range: when the load fluctuation rate exceeds the set threshold, the number of intervals is increased, and when the load is stable, the number of intervals is maintained.

[0020] The pollution characteristic table is stored in a cloud database and includes a two-dimensional index structure. The first dimension index is a pollutant type code, and the second dimension index is a concentration threshold classification table. When a composite pollutant is identified in the received operating parameters, the system activates a pollutant combination query mode. In this mode, the table calls a pollutant synergistic effect matrix, which pre-records the kinetic characteristics of different pollutant combinations interacting at specific concentration ratios. The pollution characteristic curve is characterized by a curve equation parameter table, including sixteen characteristic values such as pollutant diffusion attenuation coefficient, reaction rate constant, and saturation critical point. Each curve corresponds to a unique characteristic curve identification code.

[0021] The pollution index allocation coefficient calculation model adopts a three-layer calculation architecture. The first layer architecture processes the normalization of basic parameters, converting different dimension operating parameters into standard pollution equivalent values. The second layer architecture performs interval positioning, determining the position of the current pollution equivalent value in the parameter interval according to the maximum pollution load total span and the current parameter interval width. The third layer architecture activates the pollution characteristic curve matching, retrieving the corresponding curve equation through the characteristic curve identification code. The core of the calculation model includes a characteristic value weight allocation module, which has an environmental impact factor evaluation table built-in, where the weight coefficient of heavy metal pollutants is higher than that of organic pollutants, and the decay coefficient correction value of persistent pollutants is higher than that of degradable pollutants. The pollution index allocation coefficient is obtained by multiplying the characteristic curve equation and the weight coefficient. Each parameter interval finally outputs a floating-point allocation coefficient value.

[0022] The monitoring point quantity calculation adopts a proportional allocation method. The system presets the total number of basic monitoring points in the entire area as a constant value. The number of monitoring points in a specific parameter interval is determined by the percentage of the pollution index allocation coefficient of that interval in the sum of all interval allocation coefficients. The specific calculation formula is: interval monitoring point number = total number of basic monitoring points × (allocation coefficient of this interval / sum of allocation coefficients of all intervals). This calculation is automatically triggered after each parameter interval division, and the calculation result is rounded. When the allocation coefficient increment of a parameter interval exceeds twice the standard deviation of the historical mean value, the emergency monitoring mechanism is triggered, and the number of monitoring points in that interval is temporarily increased by a preset number of emergency points.

[0023] The deployment of key monitoring points follows the principle of spatial equal density. In the rectangular processing pool scenario, the system rasterizes the physical area corresponding to each parameter interval, and the grid side length is inversely proportional to the number of monitoring points. A key monitoring point is deployed at the center of each grid. For irregular areas, the Delaunay triangular mesh generation technology is used to generate a Delaunay triangular mesh with monitoring points as vertices, ensuring the spatial uniformity of monitoring point distribution. The monitoring point coordinate information is synchronized in real time to the dynamic monitoring focus layer, and the position deviation tolerance is controlled within five centimeters.

[0024] The pollution index analysis layer establishes a double-effect verification mechanism to ensure data reliability. The distribution coefficient generated by the initial calculation needs to be verified by comparing with the historical database. When the distribution coefficient of a specific parameter interval deviates from the historical mean value of the interval by three standard deviations, the manual review process is triggered. The system generates a three-dimensional visualization report containing a pollution parameter scatter plot, a distribution coefficient calculation tree, and an interval division heat map. The operator can manually adjust the parameter interval boundaries through the interactive interface. The confirmed parameter division scheme automatically updates the pollution characteristic table weight library, strengthening the calculation adaptability of similar pollution scenarios. When the entire analysis process is complete, the system outputs a pollution index distribution map with a timestamp, which is transmitted to the aeration path generation layer through a high-speed data bus.

[0025] The data flow closed-loop management adopts a bidirectional verification mode. Multi-source sensing terminals upload raw parameter snapshots every thirty seconds. After the analysis layer generates the monitoring point deployment scheme, it randomly selects ten percent of the monitoring point deployment coordinates for reverse verification. During verification, test instructions are sent to the specified coordinates, and the sensing terminal returns the actual monitoring value of the current coordinates. If the deviation between the theoretical parameter value corresponding to the deployment coordinates and the actual monitoring value is greater than the preset error upper limit, the system automatically marks the area as an abnormal grid and re-executes the parameter interval division calculation. Abnormal grid data is stored in an independent log file, triggering the offline training process of the deep learning model.

[0026] The system configures a three-level pollution early warning response strategy. Level one response is triggered when the distribution coefficient of a single parameter interval exceeds the safety threshold, automatically notifying the on-site management personnel. Level two response is activated when there is an abnormal increase in interval distribution coefficient for three consecutive calculation periods, forcing an increase in monitoring frequency in the abnormal area. Level three response is triggered when multiple parameter intervals exceed the standard simultaneously, directly linking to the aeration parameter generation layer to execute the emergency aeration scheme. Each response corresponds to an independent message notification template and device control instruction set, and the response level priority is dynamically increased with the pollution index.

[0027] The maintenance cycle setting adopts an adaptive algorithm. The analysis layer runs a device health evaluation program in the background, calculates the system maintenance requirement index based on twelve performance indicators such as data processing delay rate, memory peak, and network transmission packet loss rate. When the index exceeds the critical value, the hardware inspection list is pushed to the operation and maintenance platform, marking the specific device number that needs maintenance. All operation records are distributed stored by the blockchain node, forming an unalterable analysis process audit trail. After each version upgrade of the analysis layer, a forty-eight-hour gray calculation is automatically performed, and the analysis results of the new and old versions are compared. If the difference rate exceeds the limit, it is rolled back to the previous stable version.

[0028] Example 2: see Figure 3After the parameter interval division and the pollution index allocation coefficient calculation of the pollution index analysis layer, the aeration path generation layer starts to construct the aeration intensity hierarchical path. The system first receives the pollution index allocation coefficient set of all key monitoring points from the pollution index analysis layer, which is stored according to the monitoring point coordinate position index. The pollution state of each monitoring point is classified by a three-color identification system: the red identification represents the heavy pollution monitoring point with the highest pollution index allocation coefficient, the green identification represents the light pollution monitoring point with the lowest pollution index allocation coefficient, and the yellow identification represents the intermediate value monitoring point. The initial monitoring point is automatically selected by the system according to the structural characteristics of the wastewater treatment facility, and is usually located at the midpoint position of the water inlet and the aeration equipment connection line.

[0029] The pollution value difference comparison adopts the relative change rate algorithm. The pollution value of the heavy pollution monitoring point is denoted as , the pollution value of the light pollution monitoring point is denoted as , and the pollution value of the initial monitoring point is denoted as . The difference comparison formula is:

[0030] wherein: is the relative change rate difference, which is used to determine the generation order of the aeration path. When is greater than zero, the system determines that the pollution difference between the heavy pollution monitoring point and the initial monitoring point is more significant; when is less than or equal to zero, it is determined that the difference between the light pollution monitoring point and the initial monitoring point is more significant. The denominator of the formula adopts the maximum value normalization processing to eliminate the influence of the different pollution concentration dimensions.

[0031] The path generation engine includes three-level processing modules. The first-level module processes the path direction decision, and selects the priority path endpoint according to the positive and negative of the value. The second-level module calculates the path turning point, and establishes a polar coordinate system with the initial monitoring point as the origin in the two-dimensional plane coordinates of the wastewater treatment facility, and decomposes the aeration path into radial movement and tangential movement two components. The third-level module generates the path intensity curve, and the curve slope is proportional to the pollution index allocation coefficient of the endpoint monitoring point. The path generation automatically avoids the physical obstacles in the wastewater treatment facility, and the obstacle information comes from the facility three-dimensional model database.

[0032] The aeration intensity hierarchy adopts a dynamic interval division method. The system presets the basic aeration intensity as three levels: high intensity corresponds to the red monitoring point, medium intensity corresponds to the yellow monitoring point, and low intensity corresponds to the green monitoring point. The actual classification adjusts the specific value of the monitoring point pollution index, and the adjustment amplitude does not exceed twenty percent of the basic intensity. The aeration intensity change of each path segment adopts a ramp function transition to avoid mechanical impact of the aeration equipment. The aeration intensity at the path turning point is the weighted average value of the adjacent path intensities, and the weight is determined by the turning angle.

[0033] The path integration algorithm adopts spatiotemporal constraints optimization. The first aeration path is the main path, and its aeration intensity change rate is set to a standard value. The second aeration path is the return path, and the intensity decay coefficient is set to one and a half times that of the first path. The third aeration path is the auxiliary path, and the intensity fluctuation range is controlled within 60% of the main path. The three paths adopt an interleaved starting strategy in the time dimension, with the main path starting first, the return path starting five seconds later, and the auxiliary path starting ten seconds later. In the spatial dimension, the minimum distance between any two paths is ensured to be greater than the influence radius of the aeration equipment.

[0034] The aeration equipment control parameter converter converts the path information into execution instructions. The path coordinate point sequence generates a smooth curve through cubic spline interpolation, and the distance between the interpolated coordinate points is fixed at ten centimeters. Each interpolation point is associated with three control parameters: aeration fan speed percentage, aeration duration, and delay interval. The speed percentage is determined by the intensity level of the path segment where the point is located, and pulse width modulation technology is used to achieve stepless speed regulation. The aeration duration has a base value of two seconds, which is dynamically adjusted according to the pollutant degradation rate in the area where the path point is located. The delay interval ensures that the action times of adjacent aeration points do not overlap, with a minimum interval of zero point five seconds.

[0035] The path verification mechanism includes real-time feedback correction. During the execution of the aeration equipment, the three closest monitoring points form a verification triangle group, and the real-time monitoring of the dissolved oxygen concentration rate is performed. When the deviation between the actual rate and the expected rate exceeds the threshold value, the system automatically inserts a correction path point. The aeration intensity of the new path point is adjusted in proportion to the deviation, and an intensity compensation gradient is formed in the subsequent three path points. The correction record is stored in the path correction log for optimizing the subsequent path generation algorithm. After each path execution is completed, an execution report containing the actual path trajectory, intensity distribution curve, and correction point position is generated.

[0036] The abnormal handling module designs special coping strategies for four typical working conditions. The first is the path interruption condition, when the aeration equipment stops unexpectedly during path execution, the system records the interruption point coordinates and starts the adjacent standby equipment. The second is the monitoring point failure condition, when the path endpoint monitoring point data is abnormal, the system automatically switches to the alternative monitoring point and recalculates the path. The third is the water quality mutation condition, when a sudden increase in pollutant concentration occurs during path execution, the current path is immediately stopped and an emergency aeration mode is started. The fourth is the equipment conflict condition, when multiple aeration paths may interfere with each other in time and space, the system automatically calculates the optimal avoidance scheme.

[0037] The historical path database adopts a time series graph structure for storage. Each generated aeration path is encoded as a weighted directed graph, with nodes representing key points on the path and edges representing path segments and their intensity attributes. The graph is stored with a twelve-dimensional feature vector, including path total length, average intensity, turning angle variance, and other features. The database supports similar working condition retrieval functions, and when the current pollution distribution pattern is detected to have a similarity to historical records exceeding a threshold, the historical optimal path solution is preferentially called. The path optimization engine performs offline evolutionary calculations on all historical paths every week, eliminating inefficient paths and generating mutated path solutions.

[0038] The dynamic load balancing module manages multi-path parallel execution. When the sewage treatment facility needs to execute multiple aeration paths simultaneously, the system allocates path tasks based on the real-time working state of the aeration equipment group. The allocation strategy considers factors such as equipment remaining life, current load rate, and energy consumption efficiency, and uses a greedy algorithm to select the optimal equipment combination. A safety buffer zone is set between parallel paths, with the buffer zone size being proportional to the product of path intensity. The load balancing state is evaluated every thirty seconds, and when the overall load rate of the equipment group exceeds the warning line, the intensity level of non-critical paths is automatically reduced.

[0039] The aeration path generation layer is deeply integrated with the equipment maintenance system. Before each path execution, the system checks maintenance indicators such as the cumulative working time of the target aeration equipment, the latest maintenance record, and the performance decay coefficient. For equipment approaching the maintenance period, the intensity requirement of the assigned path is automatically reduced. After path execution is completed, the equipment working state database is updated, and the cumulative running time is accurate to the second. When a certain equipment is assigned a high-intensity path for three consecutive times, a preventive maintenance reminder is triggered, suggesting early maintenance. Maintenance records and path generation logs are cross-indexed to form a complete equipment life cycle management chain.

[0040] Embodiment 3: refer to Figure 4 The process parameter matching layer continuously receives the latest operating parameters from the pollution parameter sensing layer, including three core process indicators: dissolved oxygen, biochemical oxygen demand, and suspended solids concentration. The preset process parameter threshold is dynamically updated according to the design specifications of the sewage treatment plant, and each indicator contains a safety operating interval consisting of an upper threshold and a lower threshold. The absolute deviation of the current process parameters from the corresponding threshold is obtained through real-time calculation, and the calculation formula is:

[0041] wherein: represents the weighted sum of process deviations, is the measured value of the current dissolved oxygen concentration, is the dissolved oxygen concentration threshold, is the measured value of the current biochemical oxygen demand, is the biochemical oxygen demand threshold, is the measured value of the current suspended solids concentration, is the threshold value of suspended solids concentration. The weight coefficient , respectively, are the sensitivity coefficients of the three pollutants to the environment. The weight library is pre-set The reference ratio can be automatically adjusted according to seasonal parameters.

[0042] Process tolerance threshold Determined by a double-layer mechanism. The basic threshold Taken from the design documents of the sewage treatment facility, and the dynamic correction value Proportionally adjusted according to the inflow fluctuation rate. When the instantaneous flow exceeds 20% of the average flow, Expand the basic threshold by 15%; when the flow is less than 30% of the average flow, Reduce the basic threshold by 10%. The final process tolerance threshold When , the system determines that the spatial area corresponding to the parameter set has a process deviation.

[0043] The process deviation parameter set marker uses bitmap indexing technology. The sewage treatment facility plane is divided into a virtual grid of 1m x 1m, and each grid is associated with an independent parameter state register. The upper eight bits of the register store the dissolved oxygen state code, the middle eight bits store the biochemical oxygen demand state code, and the lower eight bits store the suspended solids state code. When a certain parameter triggers the condition, the corresponding state code is set to an abnormal flag. When the three abnormal flags appear at the same time, the area where the grid is located is marked as the process deviation parameter set coverage area. The register state is refreshed every five seconds, and the grid that maintains an abnormal state for three consecutive refresh periods is confirmed as a valid process deviation area.

[0044] The dynamic monitoring focusing layer includes a monitoring density conversion engine. Key monitoring points establish a mapping relationship with the virtual grid according to the coordinate distribution generated by the pollution index analysis layer. Each monitoring point carries a density classification identifier: reference level (D0), enhanced level (D1), and simplified level (D2). When the grid where the monitoring point is located is marked as the process deviation parameter set coverage area, the system performs the following operations: upgrade the density identifier of the point to D1 level; locate all monitoring points within a three-meter range around it; send a density upgrade instruction to the sensing terminal of the target monitoring point. The upgrade instruction triggers a three-layer effect: the sampling frequency is adjusted from D0 level ten minutes per time to three minutes per time; the data sampling mode is changed from single sampling to continuous three times sampling to take the average value; the sensing terminal power supply mode is switched from energy saving mode to high performance mode.

[0045] The processing of the non-process deviation area adopts a hierarchical dimension reduction strategy. The monitoring points not covered by the process deviation parameter set are automatically downgraded to D2 level. The D2 level monitoring points perform a simplified monitoring procedure: the acquisition frequency is extended to thirty minutes each time; the sampling mode is changed to single instantaneous sampling; and the sensor terminal only activates the core sensor module. For a grid area where no parameter anomaly has occurred for twelve consecutive periods, the system starts a deep sleep program. This program retains the monitoring point coordinate information but suspends actual sampling, and only estimates the parameter value through spatial interpolation of adjacent monitoring point data. When abnormal fluctuations occur in the surrounding area, the deep sleep point can restore full-function monitoring within zero point five seconds.

[0046] The dense monitoring area management adopts a clustering coordination mechanism. All D1 level monitoring points are clustered according to physical location to form monitoring clusters with a radius of no more than five meters. Each cluster elects a master node to take charge of coordination, and the master node automatically matches the sampling protocol according to the pollution characteristics within the cluster. For clusters dominated by organic pollutants, the master node instructs the member nodes to focus on collecting BOD and COD parameters; for inorganic pollution clusters, the focus is on collecting heavy metal ions and SS parameters. Cluster data transmission uses direct device communication technology, and a star-shaped topology network is established between nodes. Sampling data is uploaded to the cloud after preliminary fusion at the master node. This structure reduces network traffic by sixty percent.

[0047] The sparse monitoring area optimization adopts a mobile sensing strategy. In the D2 level monitoring point distribution area, the system deploys a mobile verification sensor. The device automatically patrols along the preset inspection route, which covers all key coordinate points in the sparse area. When the verification sensor arrives at the target area, it wakes up the nearby dormant fixed monitoring points to perform synchronous sampling. The fixed monitoring points cross-verify the sampling results with the measured values of the mobile sensor, and the monitoring points with a deviation of more than ten percent are marked as suspicious nodes. Fixed monitoring points that have been continuously verified for three times trigger the position calibration program, and the device maintenance personnel receive a maintenance work order containing GPS offset data.

[0048] The monitoring mode conversion process sets a state buffer isolation area. When a region changes from a non-deviation area to a process deviation area, the system generates a two-meter wide transition zone outside the region. The monitoring points in the transition zone maintain D0 level density, but the sampling frequency is increased to five minutes each time. This buffer design avoids data discontinuity caused by sudden changes in monitoring density. Correspondingly, when the process deviation disappears, the original dense monitoring area maintains D1 level monitoring density for six hours but cancels the cluster coordination mechanism. After the observation period ends without recurrence of abnormalities, the region is gradually downgraded to D0 level and eventually enters the D2 level state.

[0049] The dynamic strategy database stores the monitoring mode configuration parameters. The database contains twelve monitoring scene templates: high nitrogen pollution scene strengthens ammonia nitrogen monitoring frequency, and high salt wastewater scene increases conductivity monitoring dimension. The scene recognizer analyzes the pollution characteristics of the process deviation parameter set. When the proportion of abnormal ammonia nitrogen records exceeds 50%, the high nitrogen scene template is automatically matched. After the template is applied, the monitoring point density grading standard is reconfigured: the ammonia nitrogen collection frequency in the D1 level area is increased to once every minute, and the SS monitoring frequency is reduced. Template switching records are saved to the version history library, and the impact range is evaluated every time the template is updated to avoid system shock caused by frequent switching.

[0050] The resource allocation monitoring module implements dynamic load balancing. The system real-time statistics D1 level monitoring point number accounts for the proportion of the total point number, when the proportion exceeds 30%, start resource optimization program. The program calculates the optimal transmission compression ratio according to the network bandwidth utilization rate, edge computing node load, cloud platform processing queue depth three indicators. The monitoring data is changed from original transmission to differential compression transmission, and the compression algorithm can compress repeated data by up to 10:1. When the server load continues to exceed the warning line, the system automatically activates the data grading mechanism: the core process parameters maintain real-time transmission, and the secondary parameters adopt the local cache batch upload mode.

[0051] Embodiment 4: refer to Figure 5 The cross-modal mapping layer receives the dense monitoring area distribution matrix output by the dynamic monitoring focusing layer, which identifies the No. 3 area on the east side of the sewage treatment tank as the current dense monitoring area, with a coordinate range of (X12-Y18). The aeration intensity grading path data packet generated by the aeration path generation layer is also received synchronously, and the path number PATH007 contains twenty-three path nodes. The time stamp alignment program is started by the space-time correlation engine, and the dissolved oxygen change curve of the No. 12 monitoring point (coordinate X15-Y16) in the No. 3 area in the last fifteen minutes is extracted from the water quality monitoring database, with a time stamp sequence of T0 (13:00:00), T1 (13:05:30), and T2 (13:11:15). PATH007 path contains three key behavior points in the same time period: aeration equipment A starts running at T0, equipment B joins in cooperation at T1, and equipment C takes over at T2. The system binds the water quality sampling data at T1 time and the start instruction of equipment B as a synchronous event group.

[0052] The water quality parameter fluctuation feature analyzer extracts the parameter change pattern of the 12th monitoring point within the time window. The dissolved oxygen concentration rises from 1.2 mg / L to 2.4 mg / L in the T0-T1 period, with an increase of 0.24 mg / L per minute; the biochemical oxygen demand decreases from 85 mg / L to 78 mg / L, with a decrease rate of 1.4 mg / L / min; the suspended solids concentration fluctuates by ±5 mg / L. This feature is encoded as a "F013" type fluctuation feature code, containing three core attributes: dissolved oxygen rising gradient value, biochemical oxygen demand decreasing persistence, and suspended solids coefficient of variation. The operating parameters of the PATH007 path recording device B at the same time are: aeration intensity level V (corresponding to fan speed 65%), single-point aeration duration 180 seconds, and bubble diameter 0.8 mm. The system establishes a mapping relationship between the feature code F013 and the aeration intensity level V and stores it in the association mapping table.

[0053] The device scheduling decision layer scans the association mapping table and identifies that the aeration intensity of the 12th monitoring point continuously meets the standard in the T1-T2 period. This period contains four consecutive sampling periods, and the dissolved oxygen concentration in each period is maintained above the 2.0 mg / L threshold, with a total compliance duration of 625 seconds. The system combines these four consecutive compliance nodes into one aeration efficient running interval, marked as EFF_ZONE08. The efficient interval triggers device scheduling priority rearrangement: the currently working device B priority is promoted to the first place, and the subsequent path node devices are sorted by compliance duration, forming the device scheduling sequence DEV_SEQ: B(625s)-C(480s)-A(320s). The 480 seconds of device C in the sequence is derived from its historical compliance record at the adjacent 11th monitoring point.

[0054] The aeration parameter generation layer converts the EFF_ZONE08 interval parameters. The aeration fan speed instruction is generated based on the efficient interval features in three-level steps: maintain the speed at 65% in the initial stage (first 200 seconds), increase to 70% in the middle stage (200-500 seconds), and decrease to 60% in the final stage (last 125 seconds). The air blower start-stop sequence is determined according to the DEV_SEQ sorting result: start device B first and run for 625 seconds, delay for 90 seconds and then start device C for 480 seconds, and device A is on standby as a backup device. The final aeration control parameter package contains the instruction pair.

[0055] Table 1: Aeration control parameters are as follows The parameter execution verification mechanism is activated after the device starts. The dissolved oxygen rising slope of monitoring point No. 12 is monitored in real time. When the measured slope deviates from the F013 characteristic reference value by 15%, the dynamic compensation program is triggered. The compensation logic includes three levels of response: the first level of response fine-tunes the rotating speed by ±3%, the second level of response prolongs the current step length by 10%, and the third level of response enables the backup device A to cooperate. At T1+300 seconds, the dissolved oxygen increase is monitored to decrease by 12%, and the system automatically triggers the first level of response to increase the rotating speed of device B to 68%. The measured data in the 300-400 second period after adjustment returns to the reference range, and the compensation process is recorded in the execution log.

[0056] The aeration device cooperation management module handles device resource conflicts. The PATH007 path plans a new path node at coordinates X14-Y17, which requires simultaneous activation of devices B and C. The system detects that device B is already running at X15-Y16, calculates the distance between the two coordinates as 3.2 meters (greater than the minimum safety distance of 2.5 meters), and determines that parallel operation is possible. Device C receives a cooperation instruction package when it starts, dynamically adjusts its aeration angle to deflect by 15 degrees, and avoids air flow interference. In the last 60 seconds before device B stops, the system preheats the device C fan to standby speed, achieving seamless switching between operating intervals.

[0057] The historical pattern matching engine archives the current scenario. The system extracts the feature vector of the EFF_ZONE08 interval (including sixteen parameters such as influent COD load, water temperature 24°C, and PH value 7.2), and performs similarity matching with the historical case library. When similar working conditions occur in the future (similarity > 85%), the system can directly call the CMD_20240718_B001 instruction template, and only needs to adjust the operation time length coefficient according to the real-time flow. The template optimizer re-evaluates the validity period of historical instructions every month, and templates that exceed three months need to be verified by water quality improvement before they can be reused.

[0058] Example 5: The aeration effect backtracking layer starts running process monitoring after the aeration parameter generation layer issues executable control instructions. The system locks the key area range of the control instruction effect, which covers the key monitoring points within a radius of four meters centered on the aeration device coordinates. Three batches of data are collected within the first three hundred seconds before aeration, with one hundred seconds between each batch. The collected indicators include dissolved oxygen concentration, ammonia nitrogen content, and oxidation-reduction potential, which are three core parameters. The reference water quality parameters are calculated by taking the arithmetic mean of the three samples and marking the timestamp, and are stored in the dedicated reference data area. The reference water quality parameter library and the real-time dynamic monitoring flow form a double-channel comparison architecture.

[0059] After the aeration control parameters are executed, the countdown observation window is opened, and the total observation time is equal to one and a half times the set running time of the aeration equipment. In the observation window, data of the target monitoring point is collected every 40 seconds, and the real-time water quality parameters are processed by using the moving average algorithm. The reference dissolved oxygen record value of a certain coordinate point X20-Y15 is 2.1 mg / L, and the data of the three consecutive periods after aeration start are 2.4 mg / L, 2.6 mg / L, and 2.8 mg / L, and the moving average value is 2.6 mg / L. The system calculates the absolute change value by subtracting the reference value from the real-time moving average value, and the absolute change value of the dissolved oxygen of this point is +0.5 mg / L. The expected improvement threshold is dynamically set according to the improvement coefficient table of the corresponding pollutants in the “Urban Wastewater Treatment Plant Operation Standard”, and the current expected threshold of dissolved oxygen is +0.7 mg / L.

[0060] The comparator analysis finds that the absolute change value of dissolved oxygen 0.5 mg / L is less than the expected threshold 0.7 mg / L, and the system immediately activates the monitoring point capacity increasing program. Query the parameter interval code Z09 to which the coordinate point belongs, and send a capacity increasing request to the pollution index analysis layer. The capacity increasing operation includes three implementation steps: retrieving the existing number of monitoring points in the Z09 interval, the reference number is six; adding two mobile monitoring points, the position algorithm uses the triangulation method to distribute them at the edge of the interval; the new points are configured with high-precision dissolved oxygen sensors, and the sampling frequency is increased to three times per minute. At the same time, the ammonia nitrogen parameter shows a change value exceeding 40% of the expected threshold, triggering the monitoring point simplification process, and the number of monitoring points in the adjacent Z08 interval is reduced from seven to four, and the device activation permissions of edge monitoring point numbers M21, M22, and M25 are cancelled.

[0061] The update of the pollution characteristic table adopts an incremental revision strategy. The pollution characteristic curve identification code CT207 of the coordinate point X20-Y15 is extracted by the backtracking layer, which describes the correlation model between ammonia nitrogen degradation characteristics and aeration intensity. The system obtains the ammonia nitrogen concentration curve form before and after aeration: the aeration before curve slope attenuation rate is 0.15 mg / L·min, and the aeration after optimization is 0.24 mg / L·min. Accordingly, the weight coefficient of CT207 curve is adjusted by 8% of the original value. At the same time, by comparing the redox potential change curve in the same area, it is found that the actual improvement degree is lower than the predicted value, and the weight coefficient of the corresponding curve CT209 is adjusted by 6%. The weight coefficient adjustment adopts a gradient control mechanism, and the maximum correction value is limited to 10% of the original weight to avoid excessive response shock of the system.

[0062] The pollution parameter interval division logic correction module links the pollution index analysis layer. The backtracking layer outputs a parameter interval adjustment proposal, proposing three corrections for the Z09 interval: the parameter interval boundary is expanded westward by 0.5 meters to include the newly added monitoring point; the maximum pollution load upper limit of the interval is revised from 185 mg / L to 195 mg / L; and the pollution level of the interval is upgraded from B class level three to B class level two. After receiving the proposal, the analysis layer starts interval re-division calculation, and the new pollution index distribution coefficient introduces an aeration effect coefficient as a new parameter. The original distribution coefficient of a certain monitoring point is 0.82, and after adding the aeration effect coefficient 0.93, it is updated to 0.76. The re-calculation process covers all key monitoring points, generates a new version of the pollution index distribution map and marks the version number.

[0063] The data credibility verification introduces a spatiotemporal consistency test method. The system randomly selects 20% of the monitoring points for backtesting: under the same aeration parameter conditions, it runs twice and collects the water quality change curves for waveform similarity analysis. The peak position deviation of the dissolved oxygen change curve generated by three runs of a certain point is less than three seconds, and the amplitude difference rate is kept within 5% to be considered as passing the verification. The monitoring points that do not pass the verification trigger the equipment calibration process, and the sensor enters the standard solution calibration mode. After calibration, the reference parameters are re-collected. All verification records are stored in the blockchain distributed ledger after adding digital signatures, forming an unmodifiable effect backtracking certificate chain.

[0064] The abnormal fluctuation processing mechanism sets four-level response plans. The first-level response handles short-term parameter fluctuations. The dissolved oxygen of a certain point drops by 0.8 mg / L within three minutes, and the system determines that it is an equipment anomaly and immediately starts the standby aeration unit. The second-level response deals with persistent deviations. When the sampling data of a certain point does not reach the expected threshold for six consecutive times, the system sends an expert diagnosis request to the control center. The third-level response handles systematic failures. When multiple monitoring points deviate from the expected trajectory simultaneously for more than 30 minutes, the system forces the current aeration strategy to be interrupted and switches to the basic operation mode. The fourth-level response deals with equipment failures. When the feedback value of the aeration fan speed is less than 25% of the instruction value and lasts for two minutes, the system automatically pushes a device maintenance work order to the maintenance terminal.

[0065] The system maintenance interface is associated with the device life cycle management database. The effect backtracking process updates the work efficiency index of the corresponding aeration equipment each time. The efficiency index of device number B7 is initially 92, and after three aeration control operations, the average improvement effect is 87, and the updated index value is 89. When the efficiency index of a certain device decreases by more than 500% for five consecutive periods, the system marks a yellow warning state in the maintenance plan. The cross-analysis of the efficiency index and the device maintenance record generates a maintenance priority matrix to guide on-site maintenance work. During equipment downtime maintenance, related monitoring points are transferred to the baseline parameter tracking mode, continuously recording the water quality change baseline under natural conditions.

[0066] The effect backtracking data is finally archived to the historical case knowledge base. Each complete backtracking cycle forms an independent archive bag, containing four types of data assets: aeration control parameter execution package copy, water quality change curve data set, monitoring point adjustment scheme, and pollution characteristic curve correction record. The knowledge base uses an index structure based on pollution characteristics, and automatically pushes historical case references when similar water quality parameter combinations are newly detected. The knowledge base performs feature clustering analysis every quarter, merging redundant cases with a similarity of more than 90%, to improve system decision efficiency. The archiving process adds water quality fingerprint encoding, which is generated by feature value hashing of thirteen core parameters, supporting accurate retrieval functions for the entire library.

[0067] The backtracking layer and the front-end sensing device run normally in a bidirectional verification mechanism. After completing ten effect backtracking cycles, the system issues a self-checking instruction to the multi-source sensing terminal. The terminal activates the built-in verification module: the dissolved oxygen sensor is connected to a standard saturated dissolved oxygen solution, the ammonia nitrogen electrode is immersed in a standard ammonium salt solution, and the data transmission module performs full-channel error rate testing. The verification results are compared and analyzed with the backtracking data, and the devices with error rates exceeding the standard are automatically suspended from sampling qualification until they are reactivated after on-site calibration. The self-checking records and effect backtracking reports form a mutual verification chain, constituting a complete data quality guarantee system.

[0068] The feedback closed-loop delay control system adjusts the backtracking rhythm. The system dynamically calculates the typical time range for effect appearance: dissolved oxygen response delay is two to five minutes, and ammonia nitrogen response delay is eight to fifteen minutes. The starting time of backtracking operation is determined according to the current main monitoring indicators. When dissolved oxygen is the main monitoring item, the initial backtracking is started three minutes after exposure ends; when ammonia nitrogen is the main monitoring item, the time is extended to ten minutes. The delay control adopts a hierarchical buffer strategy to avoid data distortion caused by too early or too late collection. The time control parameters are updated every month based on historical data statistical analysis results to adapt to the seasonal characteristics of water quality changes.

[0069] It should be noted that, in the present document, relational terms such as first and second and the like can be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0070] While embodiments of the present application have been shown and described with reference to certain explanations, it is understood that those skilled in the art can make various changes, modifications, replacements and variations to these embodiments without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A smart monitoring and aeration control system for rural sewage based on the Internet of Things, characterized in that, include: The pollution parameter sensing layer collects the operating parameters of the sewage treatment facility in real time through multi-source sensor terminals; The pollution index analysis layer obtains the operating parameters of the wastewater treatment facilities, divides the pollution parameter ranges according to the operating parameters, and calculates the pollution index allocation coefficient. Aeration path generation layer generates aeration intensity classification paths based on pollution index allocation coefficients. The process parameter matching layer compares the current process parameters with the preset process parameter thresholds and outputs a set of process deviation parameters. Dynamic monitoring of the focusing layer dynamically adjusts the monitoring point density of the multi-source sensor terminal based on the set of process deviation parameters; The cross-modal mapping layer synchronously receives the monitoring point distribution data of the dynamic monitoring focusing layer and the aeration intensity classification path of the aeration path generation layer, and establishes a spatiotemporal correlation mapping between water quality parameters and aeration intensity. The equipment scheduling decision-making layer sorts the order of aeration equipment calls according to the spatiotemporal correlation mapping results; The aeration parameter generation layer converts the order of aeration equipment calls into executable aeration control parameters.

2. The IoT-based intelligent monitoring and aeration control system for rural sewage as described in claim 1, characterized in that, Also includes: The aeration effect feedback layer is connected to the aeration parameter generation layer. It compares the changes in water quality parameters before and after the aeration control parameters are executed, and dynamically corrects the pollution parameter interval division logic based on the changes in water quality parameters.

3. The IoT-based intelligent monitoring and aeration control system for rural sewage according to claim 2, characterized in that, The pollution index analysis layer includes: Obtain the maximum pollution load range of the wastewater treatment facility and divide the pollution load range into several parameter intervals equally. Extract the pollutant type and concentration threshold from the operating parameters, and look up the corresponding pollution characteristic curve in the pollution characteristic comparison table; The pollution index allocation coefficient for each parameter interval is obtained through the pollution index allocation coefficient calculation model. The number of monitoring points corresponding to each parameter interval is calculated based on the pollution index allocation coefficient. A number of key monitoring points are evenly distributed within the corresponding parameter range.

4. The IoT-based intelligent monitoring and aeration control system for rural sewage as described in claim 3, characterized in that, The aeration path generation layer includes: The key monitoring point with the highest pollution index allocation coefficient is marked as a heavily polluted monitoring point. The key monitoring point with the lowest pollution index allocation coefficient is marked as a light pollution monitoring point; Obtain the pollution value of the pollution characteristic curve corresponding to the initial monitoring point; Determine whether the difference between the pollution value at a heavily polluted monitoring point and the pollution value at the initial monitoring point is greater than the difference between the pollution value at a lightly polluted monitoring point and the pollution value at the initial monitoring point; If true, a first aeration path is generated from the initial monitoring point to the heavily polluted monitoring point, a second aeration path is generated from the heavily polluted monitoring point back to the initial monitoring point, and a third aeration path is generated from the initial monitoring point to the lightly polluted monitoring point. If not, a first aeration path is generated from the initial monitoring point to the lightly polluted monitoring point, a second aeration path is generated from the lightly polluted monitoring point back to the initial monitoring point, and a third aeration path is generated from the initial monitoring point to the heavily polluted monitoring point. The first aeration path, the second aeration path, and the third aeration path are integrated to form an aeration intensity classification path.

5. The IoT-based intelligent monitoring and aeration control system for rural sewage according to claim 4, characterized in that, The process parameter matching layer includes: Calculate the absolute deviation between the current process parameters and the preset process parameter thresholds; Calculate the weighted sum of dissolved oxygen deviation, biochemical oxygen demand deviation, and suspended solids deviation; When the weighted sum exceeds the process tolerance threshold, the corresponding parameter range is marked as the process deviation parameter set.

6. The IoT-based intelligent monitoring and aeration control system for rural sewage according to claim 5, characterized in that, The dynamic monitoring focusing layer includes: Identify the key monitoring points covered by the set of process deviation parameters, and upgrade the area where the key monitoring points are located into a dense monitoring area; The area covered by the non-process deviation parameter set is downgraded to a sparse monitoring area, the dense monitoring area adopts a high-frequency sampling mode, and the sparse monitoring area adopts a low-frequency sampling mode.

7. The IoT-based intelligent monitoring and aeration control system for rural sewage according to claim 6, characterized in that, The cross-modal mapping layer includes: Establish timestamp alignment between water quality parameter change curves and aeration intensity grading paths in densely monitored areas; The water quality parameter fluctuation characteristics corresponding to each node in the aeration intensity grading path are labeled, and a correlation mapping table between water quality parameter fluctuation characteristics and aeration intensity level is generated.

8. The IoT-based intelligent monitoring and aeration control system for rural sewage according to claim 7, characterized in that, The equipment scheduling decision layer includes: Extract nodes that meet the aeration intensity level from the association mapping table, sort the aeration equipment call priority according to the duration of node compliance, and merge consecutively compliant nodes into the efficient operation range of the aeration equipment.

9. The IoT-based intelligent monitoring and aeration control system for rural sewage according to claim 8, characterized in that, The aeration parameter generation layer includes: The efficient operating range of the aeration equipment is converted into a step-by-step instruction for the speed of the aeration blower. A blower start-stop sequence is generated according to the priority of the aeration equipment. The step-by-step instruction for the speed of the aeration blower and the blower start-stop sequence are combined to form executable aeration control parameters.

10. The IoT-based intelligent monitoring and aeration control system for rural sewage according to claim 9, characterized in that, The aeration effect feedback layer includes: Record the baseline water quality parameters at key monitoring points before executing the executable aeration control parameters, collect the real-time water quality parameters at the same key monitoring points after execution, and calculate the absolute change of the real-time water quality parameters relative to the baseline water quality parameters. When the absolute change does not reach the expected improvement threshold, increase the number of monitoring points in the corresponding parameter range; When the absolute change continues to exceed the expected improvement threshold, the number of monitoring points in the corresponding parameter range will be reduced. Update the weighting coefficients of the pollution characteristic curves in the pollution characteristic comparison table.

Images