A water treatment process control system based on artificial intelligence

Abstract

The invention discloses a water treatment process control system based on artificial intelligence, relates to the technical field of water pollution detection, and is used to improve the detection problem of eutrophication of a duckweed pond. The control system comprises the following steps: monitoring duckweed density and water flow information, constructing a multi-layer perceptron model based on the duckweed density and water flow information to analyze the reproduction trend of duckweed, dividing the duckweed pond into regions according to the reproduction trend of duckweed and screening out prominent eutrophication positions, detecting the dissolved oxygen content and the water body debris deposition content of the prominent eutrophication positions to analyze the eutrophication degree of each prominent eutrophication position, evaluating the overall pollution situation of the duckweed pond by using a logistic regression algorithm and calculating the pollution standard, comparing the overall pollution situation of the duckweed pond with the pollution standard, purifying the prominent eutrophication position if it exceeds the standard and then comparing it again, and selecting to save the prominent eutrophication position or adjusting the screening method to obtain more prominent eutrophication positions for purification according to the comparison result.

Description

Technical Field

[0001] The present invention relates to the technical field of water pollution detection, and more specifically, to a water treatment process control system based on artificial intelligence. Background Art

[0002] There are many water pollution detection technologies, including physical, chemical and biological detection methods. The combination of water pollution detection technology and artificial intelligence can conduct real-time monitoring and pollution assessment of polluted areas, predict pollution trends in polluted areas in advance and provide more time for treatment.

[0003] The prior art has the following deficiencies:

[0004] In the past, when dealing with eutrophic duckweed ponds, it was necessary to measure the overall duckweed density in the duckweed pond in order to predict the degree of eutrophication in the duckweed pond for pollution treatment. When the duckweed pond area is large or the overall duckweed density is difficult to measure, it will cause delays in the treatment progress and extend the treatment period. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a water treatment process control system based on artificial intelligence, which intelligently divides the duckweed pond, screens out the prominent trophication positions, evaluates the duckweed pond as a whole according to the prominent trophication positions, and performs regional targeted treatment to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A water treatment process control system based on artificial intelligence, comprising: a data monitoring module, a region division module, a pollution proofreading module and a response processing module, and signal connections between the modules;

[0008] The data monitoring module is used to monitor the density of duckweed and water flow information in the duckweed pond and transmit it to the area division module. After receiving the prominent nutrient position, it detects the dissolved oxygen content and water debris deposition amount at the corresponding position and transmits it to the pollution proofreading module;

[0009] The area division module receives the duckweed density and water flow information of each divided area in the duckweed pond, builds a multi-layer perceptron model based on the duckweed density and water flow information to analyze the duckweed reproduction trend, adjusts the divided areas according to the duckweed reproduction trend and water flow information of each area, screens out the prominent nutrient positions, and transmits them to the data monitoring module and the pollution proofreading module;

[0010] After receiving the prominent eutrophication positions, the dissolved oxygen content and the amount of water debris deposition at each position, the pollution proofreading module uses the logistic regression method to analyze the eutrophication degree of each prominent eutrophication position and evaluate the overall pollution of the duckweed pond. It compares the evaluation results with the pollution standard to determine whether the duckweed pond meets the pollution standard and sends the judgment result to the response processing module;

[0011] The response processing module is used to receive the judgment results and select to save the prominent trophication positions or purify the duckweed pond. The purification results are transmitted to the pollution proofreading module for secondary comparison. If the duckweed pond meets the pollution standard, more prominent trophication positions are obtained for purification.

[0012] In a preferred embodiment, when the data monitoring module monitors the density of duckweed in the duckweed pond, the duckweed pond is divided into N areas of equal area, and the duckweed density and water flow information of each divided area are monitored and transmitted to the area division module.

[0013] In a preferred embodiment, the water flow information is the regional water flow velocity and the regional water flow direction of each divided area. The regional division module constructs a multi-layer perceptron model through the duckweed density of each divided area and the regional water flow velocity of the corresponding area to analyze the duckweed reproduction trend, and the divided area is screened according to the duckweed reproduction trend and the screened area is determined. The screened area is expanded using the regional water flow direction of the screened area to obtain the prominent nutrient position.

[0014] In a preferred embodiment, the specific steps of the regional division module for screening the prominent trophication positions are as follows:

[0015] Screening candidate areas: Sort the candidate areas according to their duckweed density, take out the median, and randomly select n areas as candidate areas from the areas where the duckweed density exceeds the median;

[0016] Analyze the reproduction trend of duckweed: merge the duckweed density in the candidate area into a duckweed density dataset, merge the regional water flow velocity in the corresponding candidate area into a flow velocity dataset, build a multi-layer perceptron model based on the duckweed density dataset and the water flow velocity dataset, and use the output of the multi-layer perceptron model as the duckweed reproduction trend;

[0017] Screening of prominent trophication locations: Screen out candidate areas where the duckweed reproduction trend exceeds the duckweed reproduction threshold as prominent trophication locations.

[0018] In a preferred embodiment, when constructing a multi-layer perceptron model by integrating the duckweed density dataset and the water flow velocity dataset, the duckweed density dataset and the water flow velocity dataset are normalized as input data, and then the duckweed reproduction trend ratio of the candidate area is calculated by an activation function. The loss function is calculated to converge the duckweed reproduction trend ratio of the candidate area. After convergence, the duckweed reproduction trend ratio of each candidate area is output, and the candidate area is screened according to the duckweed reproduction trend ratio and the area is adjusted to obtain the prominent position of trophication.

[0019] In a preferred embodiment, the pollution proofreading module uses a logistic regression algorithm to calculate the eutrophication coefficient to evaluate the overall pollution of the duckweed pond. The evaluation process is as follows:

[0020] The pollution proofreading module first calculates the proportion of the area of ​​each prominent eutrophication location to the total area of ​​the duckweed pond, evaluates the total amount of dissolved oxygen and the total amount of water debris deposition in the duckweed pond according to the dissolved oxygen content and water debris deposition in each prominent eutrophication location and the corresponding location, and calculates the eutrophication coefficient of the duckweed pond. Then, the pollution standard is calculated using the total amount of dissolved oxygen and the total amount of water debris deposition in the conventional water pond for comparison to determine the overall pollution situation of the duckweed pond.

[0021] In a preferred embodiment, the pollution proofreading module uses a logistic regression method to calculate the eutrophication coefficient and pollution standard of the duckweed pond. When the eutrophication coefficient of the duckweed pond exceeds the pollution standard, the duckweed pond is judged to have a high pollution level; when the eutrophication coefficient of the duckweed pond is lower than the pollution standard, the duckweed pond is judged to have a low pollution level.

[0022] In a preferred embodiment, when the pollution level of the duckweed pond is low, the response processing module saves the prominent trophication positions to the system database. When the pollution level of the duckweed pond is high, the prominent trophication positions are purified and the processing time is set. When the purification treatment reaches the processing time, a detection signal is sent to the data monitoring module. The data monitoring module detects the duckweed pond and transmits the detection results to the pollution proofreading module.

[0023] In a preferred embodiment, the pollution proofreading module recalculates the eutrophication coefficient of the duckweed pond and compares it with the pollution standard. If the eutrophication coefficient of the duckweed pond still exceeds the pollution standard, the data monitoring module randomly selects n+1 divided areas as candidate areas when screening candidate areas, and repeatedly screens out more prominent eutrophication locations for purification until it is determined that the pollution level of the duckweed pond is low.

[0024] The technical effects and advantages of the water treatment process control system based on artificial intelligence of the present invention are as follows:

[0025] The present invention divides the duckweed pond into regions and screens out prominent eutrophication positions by monitoring the density of duckweed and its reproduction trend, analyzes the eutrophication degree of each prominent eutrophication position and evaluates the overall pollution of the duckweed pond by detecting the dissolved oxygen content and the amount of water debris deposition at the prominent eutrophication positions, and intentionally ignores the neutralizing effect of the lower eutrophication positions on the overall pollution of the duckweed pond, while saving the overall pollution detection cost of the duckweed pond, also improving the inclusiveness of the evaluation, striving for more processing time, comparing the overall pollution situation of the duckweed pond with the pollution standard, and purifying the prominent eutrophication positions if they exceed the standard and then comparing again, and choosing to save the prominent eutrophication positions or changing the duckweed density requirements according to the comparison results to obtain more prominent eutrophication positions for purification to meet the duckweed pond restoration standard, thereby saving a lot of manpower and material resources for duckweed pond water pollution. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a schematic diagram of a water treatment process control system based on artificial intelligence in the present invention.

[0027] Figure 2 This is a flow chart of a water treatment process control system based on artificial intelligence in the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0029] The present invention monitors the density of duckweed and water flow information, comprehensively constructs a multi-layer perceptron model based on the duckweed density and water flow information to analyze the reproduction trend of duckweed, divides the duckweed pond into regions according to the reproduction trend of duckweed and screens out prominent eutrophication positions, detects the dissolved oxygen content and the water debris deposition amount of the prominent eutrophication positions to analyze the eutrophication degree of each prominent eutrophication position, uses a logistic regression algorithm to evaluate the overall pollution situation of the duckweed pond and calculates the pollution standard, compares the overall pollution situation of the duckweed pond with the pollution standard, purifies the prominent eutrophication position if it exceeds the standard and then compares it again, and selects to save the prominent eutrophication position or adjust the screening method to obtain more prominent eutrophication positions for purification according to the comparison result.

[0030] Examples, such as Figure 1 and Figure 2 As shown, a water treatment process control system based on artificial intelligence includes: a data monitoring module, a region division module, a pollution proofreading module and a response processing module, and the modules are signal connected.

[0031] The functions of each module are as follows:

[0032] The data monitoring module is used to monitor the density of duckweed and water flow information in the duckweed pond and transmit it to the area division module. When monitoring the density of duckweed and water flow information in the duckweed pond, the duckweed pond is first divided into areas and then each divided area is monitored separately;

[0033] The area division module receives the duckweed density and water flow information of each divided area in the duckweed pond, builds a multi-layer perceptron model based on the duckweed density and water flow information to analyze the duckweed reproduction trend, adjusts the divided areas according to the duckweed reproduction trend and water flow information of each area, screens out the prominent eutrophication positions, and transmits them to the data monitoring module and the pollution proofreading module; after receiving the prominent eutrophication positions, the data monitoring module detects the dissolved oxygen content and water debris deposition amount of the corresponding positions and transmits them to the pollution proofreading module;

[0034] After receiving the prominent eutrophication positions, the dissolved oxygen content and the amount of water debris deposition at each position, the pollution proofreading module uses the logistic regression method to analyze the eutrophication degree of each prominent eutrophication position and evaluate the overall pollution of the duckweed pond. It compares the evaluation results with the pollution standard to determine whether the duckweed pond meets the pollution standard and sends the judgment result to the response processing module;

[0035] The response processing module is used to receive the judgment result and select to save the prominent eutrophication position or purify the duckweed pond, and transmit the purification result to the pollution proofreading module for secondary comparison. The pollution proofreading module analyzes whether the purified duckweed pond meets the pollution standard based on the purification result, and transmits the secondary comparison result back to the response processing module. If the standard is not met, the prominent eutrophication position is saved; if the standard is met, the prominent eutrophication position is continuously updated and purification is carried out.

[0036] It should be noted that the assessment of the overall pollution situation of the duckweed pond by screening out the prominent eutrophication locations is higher than the actual value of the overall pollution of the duckweed pond, because the neutralizing effect of the lower eutrophication locations on the overall pollution of the duckweed pond is intentionally ignored. While saving the cost of detecting the overall pollution of the duckweed pond, it also improves the inclusiveness of the assessment and gains more processing time.

[0037] The data monitoring module uses optical remote sensing technology to monitor the density of duckweed in the duckweed pond, photographs and records the overall distribution of the duckweed pond, and preliminarily divides the duckweed pond according to the recorded area of ​​the duckweed pond. The duckweed pond is divided into N areas of equal area, and the duckweed density in each divided area is monitored and the water flow information is transmitted to the regional division module.

[0038] It should be noted that the water flow information includes the regional water flow velocity and direction of each divided area. Optical remote sensing technology is a technology that uses sensors to obtain surface information from a long distance. In this case, a drone can be used to carry a shooting device. The drone can be remotely controlled to reach the sky above the duckweed pond to take pictures and record. After obtaining the overall rendering of the duckweed pond, the duckweed pond can be preliminarily divided according to its area.

[0039] After receiving the duckweed density and water flow information of each divided area in the duckweed pool, the regional division module first selects the candidate area according to the duckweed density in the divided area, and then analyzes the duckweed reproduction trend according to the duckweed density and regional water flow velocity in the candidate area, and then adjusts the size of the candidate area according to the duckweed reproduction trend, and adjusts the expansion direction of the candidate area according to the regional water flow direction, so as to select the prominent nutrient position. The specific steps are as follows:

[0040] Screening candidate areas: Sort the divided areas from small to large according to the duckweed density, take out the median, randomly select n divided areas as candidate areas from the divided areas where the duckweed density exceeds the median, and record the number of candidate areas and the total number of divided areas.

[0041] Analyze the reproduction trend of duckweed: merge the duckweed density in the candidate areas into a duckweed density dataset, merge the regional water flow velocity in the corresponding candidate areas into a flow velocity dataset, build a multi-layer perceptron model based on the duckweed density dataset and the water flow velocity dataset, and use the output of the multi-layer perceptron model as the duckweed reproduction trend.

[0042] Screening prominent eutrophication locations: Screen out candidate areas where the duckweed reproduction trend exceeds the duckweed reproduction threshold, compare the duckweed reproduction trend with the duckweed reproduction threshold, and adjust each candidate area based on the regional water flow direction to obtain the prominent eutrophication location.

[0043] It should be noted that when screening prominent eutrophication positions, the ratio of duckweed reproduction trend to duckweed reproduction threshold in each region can be used as the basis for adjusting the alternative areas. The area of ​​the alternative area can be multiplied by the ratio of the corresponding duckweed reproduction trend to the duckweed reproduction threshold to obtain the area of ​​the prominent eutrophication position, and then the expansion direction of the prominent eutrophication position can be determined according to the regional water flow direction. For example, the area of ​​the alternative area is 4 square meters, and the ratio of the corresponding duckweed reproduction trend to the duckweed reproduction threshold is 1.5, then the area of ​​the prominent eutrophication position is 6 square meters. If the regional water flow direction is east, and the original area of ​​the alternative area is a square with a side length of 2 meters, then the side parallel to the east will be expanded to 3 meters to obtain an alternative area with an area of ​​6 square meters, which will be used as the prominent eutrophication position.

[0044] In the regional division module, the duckweed density dataset and the water flow velocity dataset are integrated to build a multi-layer perceptron model to analyze the duckweed reproduction trend in each candidate area. The multi-layer perceptron model consists of three layers: input layer, hidden layer and output layer. The input layer is used to receive input features and convert the input features into easy-to-operate data. The hidden layer is used to calculate and process the converted data. The output layer is used to output the calculation results. The specific steps are as follows:

[0045] Input data: The duckweed density dataset and the water flow velocity dataset are passed into the input layer as input features. The output layer normalizes the data in the duckweed density dataset and the water flow velocity dataset and then passes them into the hidden layer.

[0046] Initialization parameters: Set the initial weights and bias parameters from the input layer to the hidden layer in the multi-layer perceptron model;

[0047] Set the hidden layer processing algorithm: The duckweed density dataset and the water flow velocity dataset are processed by setting the hidden layer activation function. The activation function is taken as an example of a relatively simple weighted function. The activation function can be constructed as follows: A = -w1a + w2b + c, where A is the output of the hidden layer, a is the normalized result of the duckweed density in the candidate area in the duckweed density dataset, b is the normalized result of the regional water flow velocity in the corresponding candidate area in the water flow velocity dataset, w1 and w2 are two initial weights set in the input layer to the hidden layer, and C is the bias parameter set in the output layer to the hidden layer;

[0048] Set the output layer processing algorithm: Set the activation function of the output layer: Where B is the output result of the output layer, is the initial duckweed reproduction threshold; it can be set to the hidden layer output result calculated by the average value of the normalized duckweed density in the duckweed density data set and the average value of the normalized regional water flow velocity in the corresponding candidate area in the water flow velocity data set;

[0049] Propagation: The multi-layer perceptron passes the input features through the input layer, hidden layer and output layer in sequence, and finally the output layer calculates the output result as the duckweed reproduction trend in each area;

[0050] Calculate duckweed reproduction threshold:

[0051] Step 1: Calculate the loss function by comparing the output of the multilayer perceptron model with the preset actual labels. The mean square error of the output results of all output layers can be calculated by the loss function: Where MSE is the mean square error, n is the number of data in the duckweed density dataset or the water flow velocity dataset, and B iThe output layer output result obtained by the output layer is the normalized result of the randomly selected i-th duckweed density in the duckweed density data set and the normalized result of the regional water flow velocity in the corresponding candidate area in the water flow velocity data set, D i is the preset i-th actual label;

[0052] Step 2: Calculate the gradient of the impact of the hidden layer and the input layer on the loss function, and take the partial derivatives of the weight and bias parameters of the hidden layer to the loss function as the gradient. The calculation formulas for the gradient of the weight and the gradient of the bias parameter are: Where d1 and d2 are the weight gradient and bias parameter gradient in the hidden layer activation function respectively;

[0053] Step 3: By setting the learning rate, adjust the weight and bias parameters through the weight gradient, bias parameter gradient and learning rate. For example, update the weight and bias parameters using the product of the learning rate and the corresponding weight gradient or bias parameter gradient: W new =W old -L*d1,c new =c old -L*d3 where W new is the adjusted weight W1 or W2, c new To adjust the bias parameters; repeat the above steps until the model converges, use the final output of the hidden layer as the duckweed reproduction threshold, and replace the initial duckweed reproduction threshold with the calculated duckweed reproduction threshold;

[0054] Output results: Output the output results of the output layer after the model converges, and use the output results of the output layer as the proportion of duckweed reproduction trend in the alternative area.

[0055] The candidate areas with regional duckweed reproduction trend ratio less than 1 are deleted, and the remaining candidate areas are calculated according to the regional duckweed reproduction trend ratio. The original area of ​​the candidate area is multiplied by the regional duckweed reproduction trend ratio to obtain a new area. According to the water flow direction of the corresponding area, the new area obtained by expanding the corresponding edge is used as the prominent position of trophication.

[0056] It should be noted that the greater the density of duckweed in the region, the more nutrients the duckweed absorbs, which is limited by the limited nutrients in the region, inhibiting the reproduction of duckweed itself; the greater the regional water flow rate, the more duckweed will spread out in the region to obtain more living area and be exposed to more nutrients, thereby promoting its own reproduction. It can be seen from the activation function of the hidden layer that the smaller the duckweed density in the candidate area or the greater the regional water flow rate, the larger the hidden layer output result, the larger the output layer output result, and the greater the regional duckweed reproduction trend. In addition, the weights and bias parameters in the hidden layer activation function are initialized by random assignments. The actual labels are used to continuously adjust the convergence weights and bias parameters to obtain the optimal weights and bias parameters to calculate the duckweed reproduction threshold, thereby improving the accuracy of the output results of the multi-layer perceptron. The actual labels are set by professionals in this field and will not be repeated here.

[0057] After receiving the prominent eutrophication positions and the dissolved oxygen content and water debris deposition amount at the corresponding positions from the data monitoring module and the regional division module, the pollution proofreading module uses the logistic regression algorithm to calculate the eutrophication coefficient to evaluate the overall pollution of the duckweed pond. The specific steps are as follows:

[0058] The pollution proofreading module first calculates the proportion of the area of ​​each prominent trophic position to the total area of ​​the duckweed pond, takes the calculated area ratio of each prominent trophic position to the total area of ​​the duckweed pond as the denominator, and takes the dissolved oxygen content and water debris deposition amount detected at the corresponding position as the numerator to estimate the total dissolved oxygen and water debris deposition amount of the duckweed pond, then sums and averages the total dissolved oxygen assessment values ​​and water debris deposition amount of each prominent trophic position respectively as the final result of the total dissolved oxygen and water debris deposition amount in the duckweed pond, and normalizes the calculated final result by dividing the total dissolved oxygen by the maximum value of the total dissolved oxygen assessment values ​​calculated for each prominent trophic position, and the same operation is performed on the total water debris deposition amount.

[0059] After normalizing the total amount of dissolved oxygen and the total amount of water debris in the duckweed pond, a logistic regression formula was constructed to calculate the eutrophication coefficient: F = 1 / 1 + e -z , where F is the eutrophication coefficient of the duckweed pond, e is the natural base, and z can be obtained by subtracting the normalized result of the total amount of water debris deposition from the normalized result of the total amount of dissolved oxygen.

[0060] After the eutrophication coefficient of the duckweed pond is calculated, the pollution proofreading module sets an area limit. Through the area limit, the geographic information system database is accessed to obtain the total dissolved oxygen and total water debris deposition of conventional ponds with an area similar to that of the duckweed pond within the area limit. The eutrophication coefficient is calculated using the same method as the pollution standard and marked as f. The overall pollution situation of the duckweed pond is judged by comparing the size relationship between the eutrophication coefficient F of the duckweed pond and f. When F>f, the pollution degree of the duckweed pond is judged to be high; otherwise, the pollution degree of the duckweed pond is judged to be low.

[0061] It should be noted that the higher the degree of eutrophication in the pond, the more duckweed there is on the water surface, which restricts the growth of aquatic plants and animals. The increase in debris leads to an increase in the amount of water debris deposited, and more bacteria are attracted to decompose it. A large amount of oxygen will be consumed during the decomposition process, resulting in a decrease in the dissolved oxygen content in the pond. From the formula for calculating the eutrophication coefficient, it can be seen that when the amount of water debris deposited in the duckweed pond is greater or the dissolved oxygen content is smaller, the calculated eutrophication coefficient is greater, and the higher the degree of eutrophication in the duckweed pond; obtaining the dissolved oxygen content and water debris deposition at the prominent eutrophication position is a conventional technical means, which will not be elaborated here. When detecting the amount of water debris deposition in the detection area, a single sediment can be selected to improve the detection efficiency. For example, the total volume of plant debris in the water is detected as the amount of water debris deposition, etc. The geographic information system database is a system for storing, analyzing and managing geographic spatial data, in which data resources in different pools can be found, including the pool area, the dissolved oxygen content in the pool, and the amount of water debris deposition. The area limit is used to limit the area of ​​the conventional pool found by the system. Its setting is not unique and can be set according to actual conditions.

[0062] The response processing module is used to perform different processing after receiving the judgment result. When the judgment result is that the pollution degree of the duckweed pond is low, the prominent eutrophication position is saved in the system database for subsequent call analysis; when the judgment result is that the pollution degree of the duckweed pond is high, the water debris sediment in the prominent eutrophication position is removed and the duckweed is broken up for aquatic animals in the pond to eat. Removing water debris can reduce oxygen consumption during the decomposition of debris, and breaking up duckweed can increase the area of ​​the water body exposed to direct sunlight, promote the absorption of nutrients by aquatic plants and animals, and thus reduce eutrophication pollution;

[0063] The response processing module sets the processing time, sends a detection signal to the data monitoring module according to the processing time, the data monitoring module detects the duckweed pond, detects the dissolved oxygen and the sediment volume of water impurities at the prominent eutrophication positions and transmits them to the pollution calibration module. The pollution calibration module calculates the eutrophication coefficient F of the duckweed pond and compares it with f. If F > f, the judgment result is sent back to the data monitoring module. When the data monitoring module selects alternative areas, randomly select n + 1 divided areas as alternative areas in the divided areas where the duckweed density exceeds the median. Repeat the operation until F < f, then the pollution degree of the duckweed pond is low, and the prominent eutrophication positions are saved to the system database.

[0064] It should be noted that to remove the sediment of water impurities or crush the duckweed, a water cleaner or a water crusher can be installed at the prominent eutrophication positions and connected to the system, and it will be turned on after making a response according to the judgment result obtained by the response processing module. The processing time is not unique and can be set according to the area size of each prominent eutrophication position in the duckweed pond, and no further analysis will be made here.

[0065] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0066] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application of the technical solution and the inventive constraints. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0067] In addition, in each embodiment of this application, the functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0068] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0069] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A water treatment process control system based on artificial intelligence, characterized in that: include: Data monitoring module, area division module, pollution proofreading module and response processing module, and signal connection between modules; The data monitoring module is used to monitor the density of duckweed and water flow information in the duckweed pond and transmit it to the area division module. After receiving the prominent nutrient position, it detects the dissolved oxygen content and water debris deposition amount at the corresponding position and transmits it to the pollution proofreading module; The area division module receives the duckweed density and water flow information of each divided area in the duckweed pond, builds a multi-layer perceptron model based on the duckweed density and water flow information to analyze the duckweed reproduction trend, adjusts the divided areas according to the duckweed reproduction trend and water flow information of each area, screens out the prominent nutrient positions, and transmits them to the data monitoring module and the pollution proofreading module; After receiving the prominent eutrophication positions, the dissolved oxygen content and the amount of water debris deposition at each position, the pollution proofreading module uses the logistic regression method to analyze the eutrophication degree of each prominent eutrophication position and evaluate the overall pollution of the duckweed pond. It compares the evaluation results with the pollution standard to determine whether the duckweed pond meets the pollution standard and sends the judgment result to the response processing module; The response processing module is used to receive the judgment results and select to save the prominent trophication positions or purify the duckweed pond. The purification results are transmitted to the pollution proofreading module for secondary comparison. If the duckweed pond meets the pollution standard, more prominent trophication positions are obtained for purification.

2. The water treatment process control system based on artificial intelligence according to claim 1 is characterized by: When the data monitoring module monitors the density of duckweed in the duckweed pond, the duckweed pond is divided into N areas of equal area, and the duckweed density of each divided area and the water flow information are respectively monitored and transmitted to the area division module.

3. The water treatment process control system based on artificial intelligence according to claim 2 is characterized by: The water flow information includes the regional water flow velocity and regional water flow direction of each divided area. The regional division module constructs a multi-layer perceptron model through the duckweed density of each divided area and the regional water flow velocity of the corresponding area to analyze the duckweed reproduction trend. The divided area is screened according to the duckweed reproduction trend and the screened area is determined. The screened area is expanded using the regional water flow direction of the screened area to obtain the prominent eutrophication position.

4. The water treatment process control system based on artificial intelligence according to claim 3 is characterized by: The specific steps for screening prominent trophication locations in the regional division module are as follows: Screening candidate areas: Sort the candidate areas according to their duckweed density, take out the median, and randomly select n areas as candidate areas from the areas where the duckweed density exceeds the median; Analyze the reproduction trend of duckweed: merge the duckweed density in the candidate area into a duckweed density dataset, merge the regional water flow velocity in the corresponding candidate area into a flow velocity dataset, build a multi-layer perceptron model based on the duckweed density dataset and the water flow velocity dataset, and use the output of the multi-layer perceptron model as the duckweed reproduction trend; Screening of prominent trophication locations: Screen out candidate areas where the duckweed reproduction trend exceeds the duckweed reproduction threshold as prominent trophication locations.

5. The water treatment process control system based on artificial intelligence according to claim 4 is characterized by: When constructing a multi-layer perceptron model based on the duckweed density dataset and the water flow velocity dataset, the duckweed density dataset and the water flow velocity dataset are used as input data for normalization, and then the duckweed reproduction trend ratio of the candidate area is calculated through the activation function. The loss function is calculated to converge the duckweed reproduction trend ratio of the candidate area. After convergence, the duckweed reproduction trend ratio of each candidate area is output. The candidate area is screened according to the duckweed reproduction trend ratio and the area is adjusted to obtain the prominent trophic position.

6. The water treatment process control system based on artificial intelligence according to claim 1 is characterized by: The pollution proofreading module uses a logistic regression algorithm to calculate the eutrophication coefficient to evaluate the overall pollution of the duckweed pond. The evaluation process is as follows: The pollution proofreading module first calculates the proportion of the area of ​​each prominent eutrophication location to the total area of ​​the duckweed pond, evaluates the total amount of dissolved oxygen and the total amount of water debris deposition in the duckweed pond according to the dissolved oxygen content and water debris deposition in each prominent eutrophication location and the corresponding location, and calculates the eutrophication coefficient of the duckweed pond. Then, the pollution standard is calculated using the total amount of dissolved oxygen and the total amount of water debris deposition in the conventional water pond for comparison to determine the overall pollution situation of the duckweed pond.

7. The water treatment process control system based on artificial intelligence according to claim 6 is characterized by: The pollution proofreading module uses the logistic regression method to calculate the eutrophication coefficient and pollution standard of the duckweed pond. When the eutrophication coefficient of the duckweed pond exceeds the pollution standard, the pollution level of the duckweed pond is judged to be high; when the eutrophication coefficient of the duckweed pond is lower than the pollution standard, the pollution level of the duckweed pond is judged to be low.

8. The water treatment process control system based on artificial intelligence according to claim 7 is characterized by: When the pollution level of the duckweed pond is low, the response processing module saves the prominent eutrophication positions to the system database. When the pollution level of the duckweed pond is high, the prominent eutrophication positions are purified and the processing time is set. When the purification treatment reaches the processing time, a detection signal is sent to the data monitoring module. The data monitoring module detects the duckweed pond and transmits the detection results to the pollution proofreading module.

9. The water treatment process control system based on artificial intelligence according to claim 8, characterized in that: The pollution proofreading module recalculates the eutrophication coefficient of the duckweed pond and compares it with the pollution standard. If the eutrophication coefficient of the duckweed pond still exceeds the pollution standard, the data monitoring module randomly selects n+1 divided areas as candidate areas when screening candidate areas, and repeatedly screens out more prominent eutrophication locations for purification until it is determined that the pollution level of the duckweed pond is low.

Images