Intelligent control system for water recycling of construction site

Abstract

The invention discloses an intelligent control system for construction site water recycling, which belongs to the field of construction site wastewater treatment and comprises a water quality monitoring module, a collection and treatment module, a chemical treatment module, a filtration treatment module, an evaluation and detection module, a clear water recovery module, an analysis and decision module, an interaction report module and a monitoring and early warning module. The system can accurately control the dosage of chemicals, reduce the operation cost, reduce the excessive use of chemicals, reduce the risk of chemical residues after sewage treatment, improve the environmental protection property, improve the accuracy of the treatment effect, improve the operation flexibility of the system, ensure the quick response of the system to the dynamic demand of a construction site, and improve the construction efficiency. Resource waste or system operation interruption caused by manual decision lag is avoided, scientificity and credibility of a system operation strategy are improved, acceptance and implementation of a construction party or a supervision party are facilitated, the comprehensive operation efficiency of the system is improved, and the service life of equipment is prolonged.

Description

Technical Field

[0001] The invention relates to the field of construction site wastewater treatment, and in particular to an intelligent control system for water recycling at a construction site. Background Art

[0002] As the global water shortage problem becomes increasingly serious, the efficient use and conservation of water resources has become the focus of global attention. As one of the high water consumption sites, construction sites have long faced huge challenges in water resource management. Traditional construction sites use water for one-time use and directly discharge it without treatment, resulting in water resource waste and environmental pollution. In response to the requirements of sustainable development, more and more construction sites have begun to adopt water recycling systems to reduce water consumption and reduce pollution to natural water bodies. However, traditional water recycling systems mostly rely on manual monitoring and regulation, which is not only inefficient, but also difficult to accurately handle the complex water quality and flow fluctuations on the construction site.

[0003] The existing intelligent control system for water recycling in construction sites cannot accurately control the dosage of chemicals, which increases operating costs. The excessive use of chemicals is more common, which increases the risk of chemical residues after sewage treatment. In addition, the existing intelligent control system for water recycling in construction sites cannot respond quickly to the dynamic needs of the construction site, and is prone to waste of resources or system operation interruption due to delayed manual decision-making, which reduces the scientific nature and credibility of the system operation strategy and is not conducive to the acceptance and implementation of the construction party or the supervisory party. Therefore, we propose an intelligent control system for water recycling in construction sites. Summary of the invention

[0004] The purpose of the present invention is to solve the defects in the prior art and to propose an intelligent control system for water recycling at a construction site.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: An intelligent control system for water recycling at a construction site, comprising a water quality monitoring module, a collection and processing module, a chemical treatment module, a filtration and processing module, an evaluation and detection module, a clean water recovery module, an analysis and decision-making module, an interactive reporting module, and a monitoring and early warning module; The water quality monitoring module is used to monitor the sewage generated during the construction process in real time; The collection and treatment module is used to centrally collect the wastewater and sewage generated during the construction process and perform preliminary treatment on the collected sewage; The chemical treatment module is used to automatically optimize the type and amount of reagents and to further remove pollutants from the pre-treated sewage; The filtering and processing module is used to dynamically adjust the filter combination and water flow rate, and filter out tiny particles and harmful components in the sewage; The evaluation and detection module includes a water quality evaluation unit and a water quality detection unit; The water quality assessment unit is used to analyze the chemical and microbiological indicators in the water body in real time and predict the short-term change trend of water quality; the water quality detection unit is used to perform trend analysis on historical data, continuously monitor the recycled water, and predict potential changes in water quality; The clean water recovery module is used to return the treated clean water to the clean water storage system after the water quality assessment unit assesses that the water quality is qualified; The analysis and decision-making module is used to form a real-time decision-making plan for system operation and dynamically optimize different decision-making plans; The interactive reporting module is used to manage information exchange between the management personnel and the system, and to generate water cycle data reports; The monitoring and early warning module is used to perform remote monitoring through the Internet of Things technology, and when an abnormality is detected, a push notification is sent to the staff via SMS or mobile application.

[0006] As a further solution of the present invention, the chemical treatment module automatically optimizes the type and amount of the reagents as follows: S1.1: Based on expert experience or historical data, weights corresponding to different evaluation criteria of reagent cost, effect and environmental friendliness are assigned. Then, the cost of minimizing the use of reagents while meeting the requirements of sewage treatment effect is set as the goal, and the corresponding objective function is set based on this goal and the weights of each group of criteria; S1.2: Each drug type and its corresponding dosage are regarded as a group of qubit codes, and the fitness value of each group of qubit codes is calculated through the objective function. The qubit code with the highest fitness value is taken as the optimal individual, and the rotation angles of the remaining qubit codes are obtained according to the preset control parameters and the fitness value of the optimal individual; S1.3: adjusting the probability amplitude of each qubit code based on the rotation angle corresponding to each qubit code, verifying whether the probability amplitude of each qubit code after adjustment meets the set normalization condition, and if not, performing normalization adjustment on it, and then updating the corresponding qubit code according to the updated probability amplitude; S1.4: After the qubit code is updated, the delivery plan is obtained by measuring the state of each qubit code, and the fitness of each new delivery plan is calculated according to the drug combination plan generated by the qubit code, and the optimal individual selection, rotation angle calculation and qubit code update are performed again; S1.5: Repeat the iteration until the fitness value of the optimal individual reaches the preset threshold, then stop, and use the type and amount of medicine corresponding to the quantum bit encoding of the optimal individual as the optimal delivery plan, and use the optimal delivery plan in the actual sewage treatment process.

[0007] As a further solution of the present invention, the specific steps of dynamically adjusting the filter combination and water flow rate of the filtering processing module are as follows: S2.1: Collect information on each filter used in the water circulation system of the construction site, and generate multiple groups of filter combinations based on the information of each group of filters. Then, generate multiple groups of filtering schemes according to each group of filter combinations and different water flow rates to construct the corresponding scheme space, and calculate the heuristic information of each group of filtering schemes based on the filtering cost and filtering effect; S2.2: Initialize the pheromone concentration of each solution in the solution space, and randomly generate multiple groups of explorers, then initialize the initial position of each group of explorers, that is, the initial filtering solution, and calculate the corresponding selection probability based on the pheromone and heuristic information of each filtering solution. According to the calculated selection probability, each explorer selects a path on the filter combination and water flow rate; S2.3: Each group of explorers builds a complete filter combination and water flow rate scheme based on the selected path, and applies the scheme to the filtration system to test its actual filtration effect and cost, and obtains the fitness of the corresponding path of the scheme. According to the filtration effect of each explorer path, the pheromone concentration on the corresponding path is updated; S2.4: Repeat the path construction and pheromone update until the pheromone change value reaches the preset threshold, and then take the path with the highest pheromone concentration as the optimal filter combination and water flow rate scheme, and use this scheme in the actual filtration process.

[0008] As a further solution of the present invention, the water quality assessment unit predicts the short-term change trend of water quality in the following specific steps: S3.1: Quantify the chemical and microbiological indicators of the water body, represent different combinations of current water quality into each different water quality state, so as to form the state space of water quality assessment, collect measures to control or adjust the water quality state, and construct the action space of water quality change; S3.2: Based on the state space and action space, calculate the probability of the current state transferring to the next state under different measures, and construct the corresponding state transfer probability matrix based on the calculated probability. Then, evaluate the expected benefit of executing any measure under the current state according to the improvement degree of pollution indicators and treatment cost; S3.3: According to the current water quality status and the path of state change, the expected benefit is iteratively calculated through the Bellman expectation equation, and then the benefits under different measures are calculated until the benefit change value converges to the preset range, and then the measure with the highest benefit is selected to control the water quality change trend.

[0009] As a further solution of the present invention, the specific steps of predicting potential changes in water quality by the water quality detection unit are as follows: S4.1: The water quality detection unit collects historical water quality data, normalizes the water quality index data of different dimensions to the same range, and divides each group of water quality index data according to a sliding window of a preset length to obtain multiple groups of time series data; S4.2: A set of change prediction models is built based on the input layer, LSTM layer and output layer architecture. The processed time series data is then divided into training set, validation set and test set. The training set data is then input into the change prediction model in sequence. After the input layer preprocesses the input data, it is transmitted to the LSTM layer. S4.3: After the input data is processed by the input gate, forget gate, candidate memory unit, update memory unit and output gate in the LSTM layer, the final hidden state is output. After repeated processing by multiple groups of LSTM layers, the hidden state generated by the last LSTM layer is processed by the output layer and the final prediction value is output; S4.4: Calculate the loss of the predicted value and the true value through the MSE loss function, propagate the loss value to each layer of the change prediction model through the chain rule, calculate the gradient of the loss value with respect to the weights and biases of each layer of the model, accumulate the gradient of each time step, obtain the overall gradient of the loss with respect to each parameter, and then use the gradient descent algorithm to update each weight and bias; S4.5: After each round of training, the performance of the change prediction model is evaluated through the validation set. If the model performance does not meet the preset expected value, retraining and validation are performed until the model performance reaches the preset expected value. The final change prediction model is output and the performance of the model on unknown data is tested through the test set. S4.6: Input the water quality data of multiple groups of past time steps selected by the staff into the change prediction model. Each group of water quality data is forward propagated through the change prediction model to output the probability of various water quality states, and output the water quality prediction value from large to small according to the state probability. It is compared with the actual observation value to detect whether there is an abnormality. If the deviation between the predicted value and the actual value exceeds the preset threshold, it is considered that the water quality has undergone an abnormal change. Combined with the probabilities of various water quality states, it is judged that the water quality indicators have a sudden change or anomaly, and the historical data characteristics that cause water quality fluctuations are marked.

[0010] As a further solution of the present invention, the specific steps of dynamically optimizing different decision-making schemes by the analysis and decision module are as follows: S5.1: The analysis and decision module assigns weights to water quality indicators, cost indicators, and treatment efficiency indicators, and constructs a comprehensive evaluation function based on each indicator and its weight to evaluate the effect of each decision plan, and randomly initializes a population of multiple decision plans, where each plan represents a potential system operation configuration; S5.2: Set multiple groups of individuals, the current position of each group of individuals represents a candidate solution, and obtain the comprehensive score of each group of decision solutions in the population through a comprehensive evaluation function. The decision solution with the highest comprehensive score is the optimal individual, and a set of coefficient vectors is set; S5.3: In each round of iteration, a random number between 0 and 1 is generated. If the generated random number is greater than the preset probability parameter, the coefficient and step length for controlling the position update are calculated based on the coefficient vector, and the remaining individuals are judged according to the size of the direction to shrink around the optimal solution; S5.4: If the coefficient is less than 1, it means that the remaining individuals surround the optimal individual, and the positions of the remaining individuals are updated according to the optimal individual, that is, the storage condition parameter values ​​of each individual are adjusted. If the coefficient is ≥ 1, it means that the remaining individuals are far away from the optimal individual. Through random search, random positions are selected in the population space, and the positions of the remaining individuals are updated. After each iteration, the coefficient vector is updated based on the linear decreasing rule. S5.5: If the generated random number is less than or equal to the preset probability parameter, the distance between the positions of the remaining individuals in the population and the position of the optimal individual is calculated, and the movement law of each individual along the spiral trajectory approaching the optimal individual is simulated through the spiral motion formula to update the individual position; S5.6: Repeat the optimal individual selection and iterative update of the position until the comprehensive score change value converges to the preset threshold range, then compare the comprehensive scores of each group of individuals, and output the plan with the highest comprehensive evaluation score as the optimal operating configuration of the system, and adjust the water recycling control in the current intelligent control system based on the optimal operating configuration combination.

[0011] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention assigns weights corresponding to different evaluation criteria of the cost, effect and environmental friendliness of the agent based on expert experience or historical data, and then takes minimizing the cost of the agent usage while meeting the requirements of the sewage treatment effect as the goal, and sets the corresponding objective function based on the goal and the weight of each group of criteria, and regards each type of agent and its corresponding dosage as a group of quantum bit codes, and calculates the fitness value of each group of quantum bit codes through the objective function, and takes the quantum bit code with the highest fitness value as the optimal individual, and obtains the rotation angle of the remaining quantum bit codes according to the preset control parameters and the fitness value of the optimal individual, and adjusts the probability amplitude of each quantum bit code based on the rotation angle corresponding to each quantum bit code, and verifies whether the probability amplitude of each quantum bit code after adjustment meets the set normalization condition, and if not, performs a normalization operation on it. Normalization adjustment is performed, and then the corresponding qubit code is updated according to the updated probability amplitude. After the qubit code is updated, the delivery plan is obtained by measuring the state of each qubit code. According to the drug combination plan generated by the qubit code, the fitness of each new delivery plan is calculated, and the optimal individual selection, rotation angle calculation and qubit code update are performed again. Repeat the iteration until the fitness value of the optimal individual reaches the preset threshold and stops. The drug type and delivery amount corresponding to the optimal individual qubit code are used as the optimal delivery plan, and the optimal delivery plan is used in the actual sewage treatment process. It can accurately control the dosage of drugs, reduce operating costs, reduce the excessive use of drugs, reduce the risk of chemical residues after sewage treatment, improve environmental protection, improve the accuracy of treatment effects, and enhance the flexibility of system operation.

[0012] 2. The present invention assigns weights to water quality indicators, cost indicators and treatment efficiency indicators, and constructs a comprehensive evaluation function based on each indicator and its weight to evaluate the effect of each decision-making plan. A population composed of multiple decision-making plans is randomly initialized, wherein each plan represents a potential system operation configuration, and multiple groups of individuals are set. The current position of each group of individuals represents a candidate plan. The comprehensive scores of each group of decision plans in the population are obtained through the comprehensive evaluation function. The decision plan with the highest comprehensive score is taken as the optimal individual. A set of coefficient vectors is set. In each round of iteration, a random number between 0 and 1 is generated. If the generated random number is greater than a preset probability parameter, the coefficient and step size for controlling the position update are calculated based on the coefficient vector. It is determined whether the remaining individuals are shrinking around the optimal solution according to the size of the direction. After each iteration, the coefficient vector is updated based on the linear decreasing rule. If the generated random number is greater than the preset probability parameter, the coefficient and step size for controlling the position update are calculated based on the coefficient vector. It is determined whether the remaining individuals are shrinking around the optimal solution according to the size of the direction. If the resulting random number is less than or equal to the preset probability parameter, the distance between the positions of the remaining individuals in the population and the optimal individual position is calculated, and the spiral motion formula is used to simulate the movement law of each individual approaching the optimal individual along the spiral trajectory, update the individual position, and repeatedly select the optimal individual and iterate the position until the change value of the comprehensive score converges to the preset threshold range. Then, the comprehensive scores of the individuals in each group are compared, and the plan with the highest comprehensive evaluation score is output as the optimal operating configuration of the system. Based on the optimal operating configuration combination, the water recycling control in the current intelligent control system is adjusted to ensure the system's rapid response to the dynamic needs of the construction site, avoid resource waste or system operation interruption caused by delayed manual decision-making, improve the scientific nature and credibility of the system operation strategy, facilitate the acceptance and implementation of the construction party or the supervisor, improve the overall operation efficiency of the system, and extend the service life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0014] Figure 1 This is a system block diagram of an intelligent control system for water recycling at a construction site proposed by the present invention. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Example 1

[0016] Reference Figure 1, an intelligent control system for water recycling at construction sites, including a water quality monitoring module, a collection and processing module, a chemical treatment module, a filtration and processing module, an evaluation and detection module, a clean water recovery module, an analysis and decision-making module, an interactive reporting module, and a monitoring and early warning module.

[0017] The water quality monitoring module is used to monitor the sewage generated during the construction process in real time; the collection and treatment module is used to centrally collect the wastewater and sewage generated during the construction process, and to perform preliminary treatment on the collected sewage; the chemical treatment module is used to automatically optimize the type and amount of chemicals, and to further remove pollutants from the pre-treated sewage.

[0018] Specifically, weights corresponding to different evaluation criteria of agent cost, effect and environmental friendliness are assigned based on expert experience or historical data. Then, the cost of minimizing the amount of agent used while meeting the requirements of sewage treatment effect is taken as the goal, and the corresponding objective function is set based on the goal and the weights of each group of criteria. Each type of agent and its corresponding dosage are regarded as a group of quantum bit codes, and the fitness value of each group of quantum bit codes is calculated through the objective function. The quantum bit code with the highest fitness value is taken as the optimal individual. The rotation angles of the remaining quantum bit codes are obtained according to the preset control parameters and the fitness value of the optimal individual, and the probability amplitude of each quantum bit code is adjusted based on the rotation angle corresponding to each quantum bit code. Verify that each Whether the probability amplitude after the adjustment of the qubit code meets the set normalization conditions, if not, it is normalized and adjusted, and then the corresponding qubit code is updated according to the updated probability amplitude. After the qubit code is updated, the delivery plan is obtained by measuring the state of each qubit code, and the drug combination plan generated by the qubit code is used to calculate the fitness of each new delivery plan, and the optimal individual selection, rotation angle calculation and qubit code update are performed again, and the iteration is repeated until the fitness value of the optimal individual reaches the preset threshold and stops, and the drug type and delivery amount corresponding to the optimal individual qubit code are used as the optimal delivery plan, and the optimal delivery plan is used in the actual sewage treatment process.

[0019] The filtration treatment module is used to dynamically adjust the filter combination and water flow rate, and filter out tiny particles and harmful components in sewage.

[0020] Specifically, the information of each filter used in the water circulation system of the construction site is collected, and multiple groups of filter combinations are generated based on the information of each group of filters. Then, multiple groups of filtering schemes are generated according to each group of filter combinations and different water flow rates to construct a corresponding scheme space. At the same time, the heuristic information of each group of filtering schemes is calculated based on the filtering cost and filtering effect, the pheromone concentration of each scheme in the scheme space is initialized, and multiple groups of exploration bodies are randomly generated. Then, the initial position of each group of exploration bodies, that is, the initial filtering scheme, is initialized, and the corresponding selection probability is calculated based on the pheromone and heuristic information of each filtering scheme. According to the calculated selection probability, each exploration body selects a path on the filter combination and water flow rate. Each group of exploration bodies constructs a complete filter combination and water flow rate scheme according to the selected path, and applies the scheme to the filtration system to test its actual filtering effect and cost, and obtains the fitness of the corresponding path of the scheme. According to the filtering effect of each exploration body path, the pheromone concentration on the corresponding path is updated, and the path construction and pheromone update are repeated until the pheromone change value is at a preset threshold and stops. Then, the path with the highest pheromone concentration is used as the optimal filter combination and water flow rate scheme, and the scheme is used in the actual filtration process. Example 2

[0021] Reference Figure 1 , an intelligent control system for water recycling at construction sites, including a water quality monitoring module, a collection and processing module, a chemical treatment module, a filtration and processing module, an evaluation and detection module, a clean water recovery module, an analysis and decision-making module, an interactive reporting module, and a monitoring and early warning module.

[0022] The evaluation and detection module includes a water quality evaluation unit and a water quality detection unit.

[0023] Among them, the water quality assessment unit is used to analyze the chemical and microbial indicators in the water body in real time and predict the short-term change trend of water quality.

[0024] Specifically, the chemical and microbial indicators of the water body are quantified, and different combinations of the current water quality are represented as each different water quality state to form a state space for water quality assessment, measures to control or adjust the water quality state are collected, and an action space for water quality change is constructed. According to the state space and the action space, the probability of the current state transferring to the next state under different measures is calculated, and the corresponding state transfer probability matrix is ​​constructed based on the calculated probability. Then, the expected benefit of executing any measure under the current state is evaluated according to the improvement degree of pollution indicators and the treatment cost. According to the current water quality state and the state change path, the expected benefit is iteratively calculated through the Bellman expectation equation, and then the benefits under different measures are calculated until the benefit change value converges to the preset range and then stops, and the measure with the highest benefit is selected to control the water quality change trend.

[0025] The water quality monitoring unit is used to perform trend analysis on historical data, continuously monitor recycled water, and predict potential changes in water quality.

[0026] Specifically, the water quality detection unit collects historical data on water quality, and normalizes the water quality index data of different dimensions to the same range, and divides each group of water quality index data according to a sliding window of preset length to obtain multiple groups of time series data. A set of change prediction models is established based on the input layer, LSTM layer and output layer architecture, and then the processed time series data is divided into training set, verification set and test set, and the training set data is input into the change prediction model in turn. After the input layer preprocesses the input data, it is transmitted to the LSTM layer. After the input data is processed by the input gate, forget gate, candidate memory unit, update memory unit and output gate in the LSTM layer, the final hidden state is output. After repeated processing by multiple groups of LSTM layers, the hidden state generated by the last LSTM layer is post-processed by the output layer, and the final predicted value is output. The loss value of the predicted value and the true value is calculated by the MSE loss function, and the loss value is propagated to each layer of the change prediction model by the chain rule. The gradient of the loss value to the weights and bias of each layer of the model is accumulated at each time step to obtain the overall gradient of the loss to each parameter, and then the gradient descent algorithm is used to update the weights and biases. After each round of training, the performance of the change prediction model is evaluated through the validation set. If the model performance does not reach the preset expected value, it is retrained and verified until the model performance reaches the preset expected value. The final change prediction model is output, and the performance of the model on unknown data is tested through the test set. The water quality data of the past multiple time steps selected by the staff are input into the change prediction model. Each group of water quality data is forward propagated through the change prediction model to output the probability of various water quality states, and the water quality prediction value is output from large to small according to the state probability, and compared with the actual observed value to detect whether there is an abnormality. If the deviation between the predicted value and the actual value exceeds the preset threshold, it is considered that the water quality has changed abnormally. Combined with the probabilities of various water quality states, it is judged that the water quality indicators have a mutation or abnormality, and the historical data characteristics that cause water quality fluctuations are marked.

[0027] The clean water recovery module is used to return the treated clean water to the clean water storage system after the water quality assessment unit assesses that the water quality is qualified; the analysis and decision-making module is used to form a real-time decision-making plan for system operation and dynamically optimize different decision-making plans.

[0028] Specifically, the analysis and decision-making module assigns weights to water quality indicators, cost indicators, and treatment efficiency indicators, and constructs a comprehensive evaluation function based on each indicator and its weight to evaluate the effect of each decision-making plan. A population composed of multiple decision-making plans is randomly initialized, where each plan represents a potential system operation configuration, and multiple groups of individuals are set. The current position of each group of individuals represents a candidate plan. The comprehensive scores of each group of decision plans in the population are obtained through the comprehensive evaluation function. The decision plan with the highest comprehensive score is the optimal individual. A set of coefficient vectors is set. In each round of iteration, a random number between 0 and 1 is generated. If the generated random number is greater than the preset probability parameter, the coefficient and step size for controlling the position update are calculated based on the coefficient vector. According to the size of the direction, it is determined whether the remaining individuals are shrinking around the optimal solution. If the coefficient is less than 1, it means that the remaining individuals are surrounding the optimal individual, and the optimal individual is selected according to the optimal individual. The positions of the remaining individuals are updated, that is, the storage condition parameter values ​​of each individual are adjusted. If the coefficient is ≥1, it means that the remaining individuals are far away from the optimal individual. Through random search, random positions are selected in the population space, and the positions of the remaining individuals are updated. After each iteration, the coefficient vector is updated based on the linear decreasing rule. If the generated random number is less than or equal to the preset probability parameter, the distance between the positions of the remaining individuals in the population and the position of the optimal individual is calculated, and the spiral motion formula is used to simulate the movement law of each individual approaching the optimal individual along the spiral trajectory, update the individual position, repeatedly select the optimal individual and iteratively update the position until the comprehensive score change value converges to the preset threshold range, then compare the comprehensive scores of each group of individuals, and output the plan with the highest comprehensive evaluation score as the optimal operating configuration of the system, and adjust the water recycling control in the current intelligent control system based on the optimal operating configuration combination.

[0029] The interactive reporting module is used to manage information exchange between managers and the system and generate water cycle data reports; the monitoring and early warning module is used for remote monitoring through the Internet of Things technology, and when an abnormality is detected, it pushes notifications to staff via SMS or mobile applications.

Claims

1. An intelligent control system for water recycling at a construction site, characterized in that: It includes water quality monitoring module, collection and processing module, chemical treatment module, filtration and processing module, evaluation and detection module, clean water recovery module, analysis and decision-making module, interactive reporting module and monitoring and early warning module; The water quality monitoring module is used to monitor the sewage generated during the construction process in real time; The collection and treatment module is used to centrally collect the wastewater and sewage generated during the construction process and perform preliminary treatment on the collected sewage; The chemical treatment module is used to automatically optimize the type and amount of reagents and to further remove pollutants from the pre-treated sewage; The filtering and processing module is used to dynamically adjust the filter combination and water flow rate, and filter out tiny particles and harmful components in the sewage; The evaluation and detection module includes a water quality evaluation unit and a water quality detection unit; The water quality assessment unit is used to analyze the chemical and microbiological indicators in the water body in real time and predict the short-term change trend of water quality; the water quality detection unit is used to perform trend analysis on historical data, continuously monitor the recycled water, and predict potential changes in water quality; The clean water recovery module is used to return the treated clean water to the clean water storage system after the water quality assessment unit assesses that the water quality is qualified; The analysis and decision-making module is used to form a real-time decision-making plan for system operation and dynamically optimize different decision-making plans; The interactive reporting module is used to manage information exchange between the management personnel and the system, and to generate water cycle data reports; The monitoring and early warning module is used to perform remote monitoring through the Internet of Things technology, and when an abnormality is detected, a push notification is sent to the staff via SMS or mobile application.

2. An intelligent control system for water recycling at a construction site according to claim 1, characterized in that: The specific steps of automatically optimizing the type and amount of reagents by the chemical treatment module are as follows: S1.1: Based on expert experience or historical data, weights corresponding to different evaluation criteria of reagent cost, effect and environmental friendliness are assigned. Then, the cost of minimizing the use of reagents while meeting the requirements of sewage treatment effect is set as the goal, and the corresponding objective function is set based on this goal and the weights of each group of criteria; S1.2: Each drug type and its corresponding dosage are regarded as a group of qubit codes, and the fitness value of each group of qubit codes is calculated through the objective function. The qubit code with the highest fitness value is taken as the optimal individual, and the rotation angles of the remaining qubit codes are obtained according to the preset control parameters and the fitness value of the optimal individual; S1.3: adjusting the probability amplitude of each qubit code based on the rotation angle corresponding to each qubit code, verifying whether the probability amplitude of each qubit code after adjustment meets the set normalization condition, and if not, performing normalization adjustment on it, and then updating the corresponding qubit code according to the updated probability amplitude; S1.4: After the qubit code is updated, the delivery plan is obtained by measuring the state of each qubit code, and the fitness of each new delivery plan is calculated according to the drug combination plan generated by the qubit code, and the optimal individual selection, rotation angle calculation and qubit code update are performed again; S1.5: Repeat the iteration until the fitness value of the optimal individual reaches the preset threshold, then stop, and use the type and amount of medicine corresponding to the quantum bit encoding of the optimal individual as the optimal delivery plan, and use the optimal delivery plan in the actual sewage treatment process.

3. An intelligent control system for water recycling at a construction site according to claim 2, characterized in that: The specific steps of dynamically adjusting the filter combination and water flow rate of the filtration processing module are as follows: S2.1: Collect information on each filter used in the water circulation system of the construction site, and generate multiple groups of filter combinations based on the information of each group of filters. Then, generate multiple groups of filtering schemes according to each group of filter combinations and different water flow rates to construct the corresponding scheme space, and calculate the heuristic information of each group of filtering schemes based on the filtering cost and filtering effect; S2.2: Initialize the pheromone concentration of each solution in the solution space, and randomly generate multiple groups of explorers, then initialize the initial position of each group of explorers, that is, the initial filtering solution, and calculate the corresponding selection probability based on the pheromone and heuristic information of each filtering solution. According to the calculated selection probability, each explorer selects a path on the filter combination and water flow rate; S2.3: Each group of explorers builds a complete filter combination and water flow rate scheme based on the selected path, and applies the scheme to the filtration system to test its actual filtration effect and cost, and obtains the fitness of the corresponding path of the scheme. According to the filtration effect of each explorer path, the pheromone concentration on the corresponding path is updated; S2.4: Repeat the path construction and pheromone update until the pheromone change value reaches the preset threshold, and then take the path with the highest pheromone concentration as the optimal filter combination and water flow rate scheme, and use this scheme in the actual filtration process.

4. The intelligent control system for water recycling at a construction site according to claim 1, characterized in that: The specific steps of the water quality assessment unit predicting the short-term change trend of water quality are as follows: S3.1: Quantify the chemical and microbiological indicators of the water body, represent different combinations of current water quality into each different water quality state, so as to form the state space of water quality assessment, collect measures to control or adjust the water quality state, and construct the action space of water quality change; S3.2: Based on the state space and action space, calculate the probability of the current state transferring to the next state under different measures, and construct the corresponding state transfer probability matrix based on the calculated probability. Then, evaluate the expected benefit of executing any measure under the current state according to the improvement degree of pollution indicators and treatment cost; S3.3: According to the current water quality status and the path of state change, the expected benefit is iteratively calculated through the Bellman expectation equation, and then the benefits under different measures are calculated until the benefit change value converges to the preset range, and then the measure with the highest benefit is selected to control the water quality change trend.

5. The intelligent control system for water recycling at a construction site according to claim 1, characterized in that: The specific steps of predicting potential changes in water quality by the water quality detection unit are as follows: S4.1: The water quality detection unit collects historical water quality data, normalizes the water quality index data of different dimensions to the same range, and divides each group of water quality index data according to a sliding window of a preset length to obtain multiple groups of time series data; S4.2: A set of change prediction models is built based on the input layer, LSTM layer and output layer architecture. The processed time series data is then divided into training set, validation set and test set. The training set data is then input into the change prediction model in sequence. After the input layer preprocesses the input data, it is transmitted to the LSTM layer. S4.3: After the input data is processed by the input gate, forget gate, candidate memory unit, update memory unit and output gate in the LSTM layer, the final hidden state is output. After repeated processing by multiple groups of LSTM layers, the hidden state generated by the last LSTM layer is processed by the output layer and the final prediction value is output; S4.4: Calculate the loss of the predicted value and the true value through the MSE loss function, propagate the loss value to each layer of the change prediction model through the chain rule, calculate the gradient of the loss value with respect to the weights and biases of each layer of the model, accumulate the gradient of each time step, obtain the overall gradient of the loss with respect to each parameter, and then use the gradient descent algorithm to update each weight and bias; S4.5: After each round of training, the performance of the change prediction model is evaluated through the validation set. If the model performance does not meet the preset expected value, retraining and validation are performed until the model performance reaches the preset expected value. The final change prediction model is output and the performance of the model on unknown data is tested through the test set. S4.6: Input the water quality data of multiple groups of past time steps selected by the staff into the change prediction model. Each group of water quality data is forward propagated through the change prediction model to output the probability of various water quality states, and output the water quality prediction value from large to small according to the state probability. It is compared with the actual observation value to detect whether there is an abnormality. If the deviation between the predicted value and the actual value exceeds the preset threshold, it is considered that the water quality has undergone an abnormal change. Combined with the probabilities of various water quality states, it is judged that the water quality indicators have a sudden change or anomaly, and the historical data characteristics that cause water quality fluctuations are marked.

6. The intelligent control system for water recycling at a construction site according to claim 1, characterized in that: The specific steps of the analysis and decision module for dynamically optimizing different decision schemes are as follows: S5.1: The analysis and decision module assigns weights to water quality indicators, cost indicators, and treatment efficiency indicators, and constructs a comprehensive evaluation function based on each indicator and its weight to evaluate the effect of each decision plan, and randomly initializes a population of multiple decision plans, where each plan represents a potential system operation configuration; S5.2: Set multiple groups of individuals, the current position of each group of individuals represents a candidate solution, and obtain the comprehensive score of each group of decision solutions in the population through a comprehensive evaluation function. The decision solution with the highest comprehensive score is the optimal individual, and a set of coefficient vectors is set; S5.3: In each round of iteration, a random number between 0 and 1 is generated. If the generated random number is greater than the preset probability parameter, the coefficient and step length for controlling the position update are calculated based on the coefficient vector, and the remaining individuals are judged according to the size of the direction to shrink around the optimal solution; S5.4: If the coefficient is less than 1, it means that the remaining individuals surround the optimal individual, and the positions of the remaining individuals are updated according to the optimal individual, that is, the storage condition parameter values ​​of each individual are adjusted. If the coefficient is ≥ 1, it means that the remaining individuals are far away from the optimal individual. Through random search, random positions are selected in the population space, and the positions of the remaining individuals are updated. After each iteration, the coefficient vector is updated based on the linear decreasing rule. S5.5: If the generated random number is less than or equal to the preset probability parameter, the distance between the positions of the remaining individuals in the population and the position of the optimal individual is calculated, and the movement law of each individual along the spiral trajectory approaching the optimal individual is simulated through the spiral motion formula to update the individual position; S5.6: Repeat the optimal individual selection and iterative update of the position until the comprehensive score change value converges to the preset threshold range, then compare the comprehensive scores of each group of individuals, and output the plan with the highest comprehensive evaluation score as the optimal operating configuration of the system, and adjust the water recycling control in the current intelligent control system based on the optimal operating configuration combination.

Images