Gate pump group flood control and drainage optimization scheduling method, equipment, medium and program product

Abstract

The invention provides a gate pump group flood control and drainage optimization scheduling method and device, a medium and a program product, and relates to the field of urban flood control and drainage. The method comprises the steps of firstly obtaining historical flood and gate pump operation data of a target area, and inputting the historical flood and gate pump operation data into a long-short-term memory neural network to construct a prediction model which is used for flood change conditions under different scheduling strategy combinations; setting different initial conditions to simulate calculation to determine a strategy combination effect, further determining flood control and drainage effects, selecting an optimal combination, and taking the optimal combination and the initial conditions as a sample training network. And then receiving real-time meteorological and hydrological data, determining an actual optimal combination and sending the actual optimal combination to a display end. According to the method, a scheduling strategy is scientifically determined through data and a model, dynamic adjustment can be carried out according to real-time data, and scientificity and effectiveness of flood control and drainage are improved.

Description

Technical Field

[0001] The present application relates to the field of urban flood control and drainage, and in particular to a flood control and drainage optimization scheduling method, equipment, medium and program product of a gate pump group. Background Art

[0002] With the acceleration of urbanization, the threat of flood disasters faced by cities is becoming increasingly serious. For target areas in cities, how to effectively carry out flood control and drainage scheduling is a key issue to ensure the safety and normal operation of the city.

[0003] In the prior art, some flood control and drainage systems collect flood data and gate pump operation data within a certain period of time, and then use simple statistical analysis methods, such as calculating statistical quantities such as average values ​​and standard deviations. Fixed scheduling strategies are formulated based on these statistical data. For example, based on the average rainfall in the past few years and the corresponding water level changes, it is set that when the water level reaches a certain threshold, a certain number of pumps are turned on or off, and the gate is adjusted to a specific opening.

[0004] However, existing technologies make it difficult to dynamically adjust scheduling strategies based on real-time meteorological and hydrological changes. Instead, they are mechanically executed according to pre-set rules, which are not very timely and have poor flexibility. Summary of the invention

[0005] The present application provides a method, device, medium and program product for optimizing the scheduling of flood control and drainage of a gate pump group, which is used to achieve reasonable scheduling of the gate pump group, effectively control the water level in the target area, and ensure urban safety and normal operation.

[0006] In the first aspect, the present application provides a flood control and drainage optimization scheduling method for a gate pump group, which is applied to a scheduling server. The method includes: obtaining historical flood data and corresponding historical gate pump operation data of a target area at multiple historical time points, wherein the historical gate pump operation data includes gate opening and pump start-stop status; inputting the historical flood data and the historical gate pump operation data into a long short-term memory neural network for processing, and constructing a flood change prediction model, wherein the flood change prediction model is used for flood changes under different scheduling strategy combinations, wherein the scheduling strategy combination includes different combinations of the gate opening and the pump start-stop status; after setting different initial conditions, multiple simulation calculations are performed through the flood change prediction model. , to determine the flood change prediction under multiple scheduling strategy combinations, the initial conditions include different rainfall conditions; determine the flood control and drainage effect of the target area under the scheduling strategy combination according to the flood change prediction, and the flood control and drainage effect at least includes the water level control condition; after determining the optimal scheduling strategy combination according to the flood control and drainage effect, use the optimal scheduling strategy combination and the corresponding initial conditions as training samples to continue training the long short-term memory neural network; receive real-time meteorological and hydrological data of the target area, extract the actual initial conditions, input the actual initial conditions into the long short-term memory neural network, and determine the actual optimal scheduling strategy combination; send the optimal scheduling strategy combination to the display end for display.

[0007] By adopting the above-mentioned technical solution, this method can make full use of the historical data mining rules, accurately predict the changes in floods under different strategies through the learning ability of neural networks, and then determine the optimal strategy. This makes flood control and drainage scheduling more scientific and reasonable, and can also adapt to weather changes in a timely manner, thereby improving the flexibility and timeliness of responding to flood disasters.

[0008] In combination with some embodiments of the first aspect, in some embodiments, after receiving real-time meteorological and hydrological data of the target area and extracting the actual initial conditions, the actual initial conditions are input into the long short-term memory neural network to determine the actual optimal scheduling strategy combination. It also includes: receiving weather warnings from the meteorological terminal of the target area; judging whether there is heavy rain within the set time in combination with the weather warning; if there is heavy rain, sending the optimal scheduling strategy combination to the display terminal in advance at a preset time; continuously tracking the information changes of the weather warning, and dynamically combining the optimal scheduling strategy.

[0009] By adopting the above technical solution and utilizing real-time weather information, we can make preparations before the rainstorm arrives, so that the gate pumps can operate in accordance with the optimal strategy in a timely manner, avoiding untimely flood control and drainage caused by sudden rainstorms. This effectively improves the ability to cope with extreme weather such as rainstorms and ensures the safety of flood control and drainage in the target area.

[0010] In combination with some embodiments of the first aspect, in some embodiments, after the step of inputting the historical flood data and the historical gate pump operation data into the long short-term memory neural network for processing and constructing a flood change prediction model, it also includes: dividing the target area into multiple sub-areas according to the vegetation coverage rate in the target area; obtaining the rainwater interception and infiltration amount of each sub-area, and optimizing the flood change prediction model in combination with the historical flood data and gate pump operation conditions corresponding to each sub-area.

[0011] By adopting the above technical solutions, areas with different vegetation coverage have different rainwater interception and infiltration. Considering these factors can make the model more accurately reflect the actual flood situation. The optimized model can better predict flood changes and improve the accuracy of scheduling strategy formulation, so as to more effectively carry out flood control and drainage and reduce the impact of flood disasters on the target area.

[0012] In combination with some embodiments of the first aspect, in some embodiments, after obtaining the rainwater interception and infiltration amount of each sub-area and combining the historical flood data and gate pump operation conditions corresponding to each sub-area, and optimizing the flood change prediction model, it also includes: obtaining multiple water quality pollution index data based on water quality monitoring equipment pre-installed in the target area; determining the polluted area based on the multiple water quality pollution index data; setting the regional weight based on the water quality pollution index data in the polluted area and the set threshold; and adjusting the optimized flood change prediction model based on the regional weight.

[0013] By adopting the above technical solutions, the scheduling strategy that takes into account water pollution and avoids sewage diffusion is given priority in the model. This allows the formulated scheduling strategy to not only prevent floods and drainage, but also protect the regional environment, prevent sewage from spreading due to improper gate pump operation, and improve the environmental performance of the entire flood control and drainage system.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining the flood control and drainage effect of the target area under the scheduling strategy combination based on the predicted flood change situation, it also includes: determining the energy utilization efficiency of the pump under each scheduling strategy combination, and the energy utilization efficiency is determined by the ratio of the pump's discharge volume to its energy consumption; adjusting the flood control and drainage effect according to the energy utilization efficiency.

[0015] By adopting the above technical solutions and considering energy utilization efficiency, it is possible to choose a more energy-saving and efficient scheduling strategy. This can not only achieve the goal of flood control and drainage, but also reduce energy consumption, improve the economy of the operation of the sluice pump group, and meet the requirements of sustainable development.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after the step of adjusting the flood control and drainage effect according to the energy utilization efficiency, it also includes: determining the total time taken for each scheduling strategy combination to achieve the expected flood control and drainage effect from the beginning of execution under different initial conditions; and adjusting the flood control and drainage effect again according to the total time.

[0017] By adopting the above technical solutions and taking time factors into account, the model can be more inclined to choose efficient scheduling strategies. This further improves the efficiency of flood control and drainage, allowing the target area to get rid of flood threats more quickly and reduce the losses caused by flood disasters.

[0018] In combination with some embodiments of the first aspect, in some embodiments, after determining the optimal scheduling strategy combination based on the flood control and drainage effect, using the optimal scheduling strategy combination and the corresponding initial conditions as training samples, and continuing to train the long short-term memory neural network, it also includes: after determining the impact of different scheduling strategy combinations on the upstream and downstream areas, evaluating different scheduling strategy combinations according to preset upstream and downstream impact evaluation indicators; if the scheduling strategy combination causes multiple indicators of the upstream and downstream areas to exceed the corresponding indicator preset thresholds, marking the corresponding scheduling strategy combination as an adverse impact strategy; when subsequently training the long short-term memory neural network, reducing the weight of the adverse impact strategy in the training sample.

[0019] By adopting the above technical solutions, after determining the optimal strategy combination and training the network, the impact on the upstream and downstream areas is evaluated and the training weights are adjusted. The impact of the upstream and downstream areas is considered to avoid over-learning of strategies with adverse effects. This allows the formulated scheduling strategy to comprehensively consider the upstream and downstream conditions, avoid adverse effects on the upstream and downstream, and ensure the rationality and effectiveness of the flood control and drainage system in the entire region.

[0020] In a second aspect, the present application provides a server, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0021] In a third aspect, the present application provides a computer-readable storage medium comprising instructions, which, when executed on a server, causes the server to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, the present application provides a computer program product, which, when executed on a server, enables the server to execute the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Due to the technical means of building a long-short-term memory neural network prediction model based on historical data and determining the optimal scheduling strategy combination through simulation calculation, the technical problem of fixed scheduling strategies in the existing technology and difficulty in dynamic adjustment according to real-time changes is effectively solved, thereby achieving a more scientific and reasonable flood control and drainage scheduling, and improving the accuracy and effectiveness of flood control and drainage effects such as water level control in the target area.

[0024] 2. Due to the technical means of dividing the area according to the vegetation coverage rate and optimizing the flood change prediction model in combination with relevant data, the technical problem that the model in the existing technology cannot accurately reflect the actual flood situation is effectively solved, thereby achieving the technical effect of improving the accuracy of scheduling strategy formulation, more effectively carrying out flood control and drainage, and reducing the impact of flood disasters on the target area.

[0025] 3. Since the technical means of calculating the energy efficiency of the pump after determining the flood control and drainage effect and adjusting the effect accordingly is adopted, the technical problem of not considering the energy utilization efficiency in the existing technology is effectively solved, thereby achieving the technical effect of reducing energy consumption while achieving the flood control and drainage goals and improving the economic efficiency of the gate pump group operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of a method for optimizing the scheduling of flood control and drainage of a gate pump group in an embodiment of the present application; Figure 2 This is another flow chart of the method for optimizing the scheduling of flood control and drainage of a gate pump group in an embodiment of the present application; Figure 3 It is a schematic diagram of a physical device structure of a scheduling server in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more listed items.

[0028] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0029] For ease of understanding, the following is a description of the process of the method provided by this implementation. Figure 1 , which is a flow chart of the optimized scheduling method for flood control and drainage of a gate pump group in an embodiment of the present application.

[0030] S101, obtaining historical flood data of a target area at multiple historical time points and corresponding historical gate pump operation data, wherein the historical gate pump operation data includes gate opening and pump start / stop status; The dispatch server needs to obtain detailed rainfall data, water level data and other flood-related data in the target area over the past years from the meteorological and hydrological departments. This data can be obtained regularly from the meteorological and hydrological database server by establishing a secure connection and using protocols such as FTP / SFTP. In addition to obtaining data from the meteorological and hydrological departments, the dispatch server can also collect public flood-related data and information in the target area over the past years through Internet crawler technology. These data can be obtained from government websites, third-party platforms, etc. The crawler program can use multi-threading, caching and other technologies to improve crawling efficiency. These crawled data can also be stored in the big data storage system. In addition, the dispatch server can monitor and record the water level data of the target area in real time through the water level monitoring equipment set up in advance. Through continuous monitoring, detailed water level information at different time periods and different flood levels in the area can be obtained.

[0031] The dispatch server needs to connect with the regional flood control department to obtain the gate pump operation logs of the target area over the past years. The logs need to include the opening and closing status of the gates at different times, the size of the opening, and the start and stop status of the pumps, speed and other operating parameters. The flood control department can store these logs in a relational database, and the dispatch server can obtain the required data through the communication connection. If the flood control department cannot provide the data directly, the dispatch server can also collect it on site. The dispatch server records the operating parameters of the gates and pumps in real time through sensors pre-installed on the gates and pump equipment.

[0032] After obtaining the above two types of data, the dispatch server needs to clean and deduplicate the data, integrate it into a relational database or big data platform, and establish a data table containing flood data at multiple historical time points in the target area and the operating parameters of the gate pump at the corresponding time. Geographical location and other information can be added to establish a spatial database to facilitate flood simulation.

[0033] S102, inputting the historical flood data and the historical gate pump operation data into a long short-term memory neural network for processing, and constructing a flood change prediction model, wherein the flood change prediction model is used for flood change conditions under different scheduling strategy combinations, wherein the scheduling strategy combination includes different combinations of the gate opening and the start / stop state of the pump; Collect flood data of the target area in the past (for example, the past 10 years), including water level, rainfall, flood range and other data. At the same time, collect historical operation data of the gates and pumps in the area, including information such as gate opening and pump start and stop status. Preprocess these historical data, including cleaning out erroneous data, removing outliers, normalization, etc., to obtain standardized data input into the long short-term memory neural network. According to the dimensions and types of input and output data, design the network structure and determine the number of nodes contained in the input layer, multiple hidden layers and output layer. The hidden layer adopts LSTM (long short-term memory) structure, which can learn the long-term dependency of time series data. The network structure design should fully consider the data characteristics. Generally, it can start with a simple structure and gradually adjust it after evaluating the results. After the network training converges, save the optimized network structure and parameters to form a prediction model for flood changes. The model accepts gate opening and pump start and stop status as input, and can output flood changes in the target area for a period of time in the future under this scheduling strategy combination, including water level, flood area, etc.

[0034] Use new flood scenarios and gate pump operation data to verify the prediction effect of the model. If the effect is not good, it is necessary to readjust the network structure, training parameters, etc. and train the network again. Repeat the optimization process until the model meets the set prediction accuracy requirements. After sufficient verification, the optimized model can be deployed to the scheduling system. When evaluating different scheduling strategies later, the strategy combination is input into the model to obtain the flood change prediction under the combination, which is used to evaluate the flood control and drainage effects of different scheduling schemes.

[0035] S103, after setting different initial conditions, performing multiple simulation calculations through the flood change prediction model to determine flood change prediction situations under multiple combinations of the scheduling strategies, wherein the initial conditions include different rainfall situations; According to historical data, different possible initial rainfall conditions are set, such as light rain, moderate rain, heavy rain, rainstorm and other different levels of rainfall. The impact of rainfall on flooding in the target area should be fully considered, including factors such as rainfall intensity, duration, and range. In addition to rainfall, other environmental conditions such as season, temperature, and topographical features of the target area should also be considered to form a complete combination of initial conditions. According to the operating range and combination mode of the gate and pump, multiple different scheduling strategy combinations are determined, and each combination represents a matching method of the gate opening and the start and stop status of the pump. Common and extreme strategy situations should be covered, such as full open / closed gates, full open / stop pumps, etc.

[0036] According to the digital elevation model, pipe network system, water body distribution and other data of the target area, a digital simulation environment is built to simulate the flow changes and waterlogging spread process under different scheduling strategies. The environment must contain key factors that affect floods. Under the set initial conditions, different scheduling strategy combinations are applied to the digital simulation environment one by one. Based on the principles of hydraulics, the numerical simulation of the flood process is calculated to obtain the water level and waterlogging point distribution of the target area under different strategies. The mapping relationship between different strategy combinations and simulation results is recorded. All the set initial rainfall conditions are looped through, and the above simulation process is repeated for each condition to obtain the flood results corresponding to all strategy combinations under the initial conditions. Finally, all simulation results under different initial conditions are summarized to form a data set of the corresponding relationship between scheduling strategies and flood changes.

[0037] S104, determining the flood control and drainage effect of the target area under the scheduling strategy combination according to the flood change prediction, and the flood control and drainage effect at least includes the water level control situation; From the simulated flood results, extract key evaluation indicators, such as the highest water level in different areas, the distribution range of waterlogging points, duration and other characteristics, water level process lines of monitoring points and important areas, etc., to accurately describe the flood process. According to the flood control standards of the target area, set specific indicators for evaluating the effects of flood control and drainage, such as the proportion of time that does not exceed a certain water level, the proportion of waterlogging area, etc. The indicators should fully consider factors such as water level control, waterlogging point elimination, and duration. Based on the above extracted flood characteristics, calculate the evaluation indicator results corresponding to each scheduling strategy combination. For example, a certain strategy makes the time proportion of the key point water level not exceed the safe water level 95%, which meets the flood control needs.

[0038] According to the index results, the flood control and drainage effects of different scheduling strategy combinations are judged. Priority is given to strategy combinations that can make the evaluation indicators meet the flood control standard requirements, such as controlling the water level and reducing the area of ​​waterlogged points. Under different initial conditions, the scheduling strategy combination with the best effect is determined on the premise of meeting the needs of flood control and drainage. These combinations can effectively play the role of gate storage and pump drainage, and realize the control of water levels in target areas and the elimination of waterlogged points. Through calculation, evaluation and comparison, the flood control and drainage effects of each scheduling strategy combination can be scientifically determined, providing a basis for the subsequent selection of the optimal scheduling strategy. Taking into full consideration various influencing factors can make the scheduling results more practical.

[0039] In some embodiments, after simulating and calculating different combinations of scheduling strategies, it is necessary to evaluate the energy efficiency under each combination scheme. Here, the energy efficiency of the pump station is mainly considered. The technical parameters of the pump station can be monitored or consulted to obtain the design value of its displacement per unit time and the corresponding design value of energy consumption. Then, when simulating and calculating different scheduling strategies, the total displacement and total energy consumption of the pump station under the strategy can be estimated according to the start and stop and operation time of the pump station determined by the strategy. Finally, the total displacement of the pump station under the strategy is divided by the total energy consumption to obtain the energy utilization efficiency of the pump station corresponding to the strategy. The higher the efficiency, the more water can be discharged per unit energy consumption, and the more reasonable the energy utilization. This efficiency index can evaluate the economy of different strategy schemes. After obtaining the energy utilization efficiency index of each scheduling strategy combination, the flood control and drainage effect of the strategy can be adjusted according to the efficiency.

[0040] The method is to add the impact of energy utilization efficiency to the flood control and drainage effect score: standardize the efficiency indicators of all strategies to a range of 0-1; multiply a weight (such as 20%) in the flood control and drainage effect score, and add the efficiency index multiplied by the weight (80%) to obtain the adjusted flood control and drainage effect. The weight is determined according to the actual situation. The higher the efficiency weight, the more obvious the adjustment. After this adjustment, the flood control and drainage effect of the less efficient strategy will be reduced, while the more efficient one will be improved. In this way, when determining the optimal strategy, the economic benefit factor can be more comprehensively considered.

[0041] In some embodiments, after simulating and calculating different scheduling strategy combinations, in addition to evaluating the flood control and drainage effects, it is also necessary to estimate the time required for strategy execution. Specifically, after setting different initial conditions (rainfall, waterlogging, etc.), simulate and run each scheduling strategy combination, and record the time from the start of strategy scheduling to the water level drop calculated by the model to achieve the expected effect during the simulation. The strategy time obtained by simulating under different initial conditions can be taken as the average or maximum value, as the total time of the strategy, and the above process can be repeated to obtain the average total time for each scheduling strategy combination to reach the expected flood control and drainage effect from the beginning under different initial conditions. This time indicator reflects the speed of strategy response and execution, and can evaluate its ability to respond to rainfall and floods. After obtaining the execution time data of each strategy combination, its flood control and drainage effect can be adjusted again according to the length of time: the time of different strategies is standardized and mapped to a range of 0-1, and the time factor is added to the effect score, for example, the standardized time is added according to a certain weight ratio. The shorter the time, the higher the score. The weight setting needs to consider the importance of the time factor, generally 10%~30% is appropriate. The flood control and drainage effect score of such a strategy combination with a short execution time will be appropriately increased, and it may be given priority when optimizing the strategy.

[0042] By considering the impact of strategy execution time on the effect, we can select the optimal scheduling strategy combination that can achieve flood control goals and has the characteristics of rapid response, thereby improving the ability to respond to major rainfall and floods.

[0043] S105, after determining the optimal scheduling strategy combination according to the flood control and drainage effect, use the optimal scheduling strategy combination and the corresponding initial conditions as training samples to continue training the long short-term memory neural network; According to the optimal scheduling strategy combination under different initial conditions determined in the previous step, these strategy combinations and the corresponding initial condition data are organized into a new sample data set for further long short-term memory neural network. Since the new sample already contains the optimal strategy results, it can be directly labeled. The best strategy label under each initial condition is given in the new sample, which provides good labeled data for neural network training using supervised learning methods. Based on the existing network model structure, the network is continued to be trained using the newly added sample data set. Unlike the initial training, a fine-tuning strategy can be adopted here, that is, freezing the network weights of the previous layers and only training the fully connected layers behind. Finally, a new version of the network model with improved prediction performance is obtained, which can replace the previous model for subsequent optimal strategy prediction work. The entire fine-tuning process enables the model to better learn the knowledge to determine the optimal strategy.

[0044] In some embodiments, after obtaining the flood control and drainage effects of different scheduling strategy combinations, it is also necessary to investigate the impact of the strategy on the upstream and downstream of the target area. Virtual hydrological and water quality monitoring points can be set up in the upstream and downstream of the target area to simulate and calculate the actual impact of different strategies on the upstream and downstream. The impact can be evaluated by indicators including: changes in upstream and downstream water levels, whether new floods occur, whether the risk of overflowing the embankment is increased, whether the water quality is deteriorated, etc. The corresponding quantitative evaluation criteria are pre-set according to each impact indicator, such as a negative impact if the water level change exceeds 0.5 meters. The upstream and downstream impact of the strategy can be judged by comparing the simulation results with the standards. After the above process evaluation, the strategy combination that has adverse effects on the upstream and downstream can be screened out. The specific judgment criteria are: if a strategy causes multiple upstream and downstream impact indicators to exceed the threshold, it is judged as an adverse impact strategy. The more indicators that exceed the threshold, the more serious the adverse impact. The weight calculation of the comprehensive super-threshold score can also be set. If the score exceeds a certain value, it is judged as an adverse impact strategy.

[0045] Mark the strategies with adverse effects, and in subsequent model training, give these strategies lower weights or filter them to avoid choosing flood control solutions that have a negative impact on the surrounding areas. After marking the strategies with adverse effects, when using these strategy samples to train the model, you need to reduce their weights: you can directly set the weights of the bad strategy samples to a lower value, such as 0.1, and the normal samples to 1. You can also add a penalty term to the model's loss function and add additional penalties to bad samples to reduce their impact on model training. You can also remove some bad samples when building the training set.

[0046] By reducing the weight of bad samples, the trained model will rarely give bad scheduling strategy suggestions in actual predictions, so as to comprehensively consider the impact on upstream and downstream and make correct flood control decisions.

[0047] S106, receiving real-time meteorological and hydrological data of the target area, extracting actual initial conditions, inputting the actual initial conditions into the long short-term memory neural network, and determining an actual optimal scheduling strategy combination; Through the meteorological and hydrological monitoring system, the real-time rainfall, water level, flow and other data of the target area are obtained to reflect the actual initial conditions at that moment. The collected real-time data is processed, including quality control, filtering, normalization, etc., to obtain the initial condition description that meets the model input requirements, mainly to characterize the current rainfall situation. The processed real-time initial conditions are input into the fine-tuned long short-term memory neural network, and the forward calculation is run. The network will predict the optimal scheduling strategy combination under the initial conditions. The network output results are analyzed and converted into a specific gate opening value and pump start and stop state combination to form the optimal scheduling strategy. According to the preset rules, such as considering the scheduling frequency limit, the recommended strategy is verified to make it meet the actual operation constraints. Finally, the optimal scheduling plan that can be directly executed is formed. When a new round of real-time data arrives, the process is repeated to achieve closed-loop feedback and continuously recommend and adjust the current optimal strategy.

[0048] In some embodiments, in order to keep abreast of the bad weather that may occur in the target area, it is necessary to establish a connection with the meteorological department in the area and receive the weather warning information they issue in real time. First, it is necessary to determine the meteorological agency that issues authoritative weather warnings in the target area. Such as the meteorological station in the province or city where the target area is located. Contact the meteorological agency to establish a data interface for receiving and parsing the warning information it issues. The warning information is generally given in the form of scheduled information release or update, and data can be obtained by subscribing to the information release channel of the meteorological agency. The received warning information needs to be parsed to extract important matters such as the warning type (such as heavy rain, hail, etc.), warning level, warning range (to determine whether it affects the target area), and warning time. After obtaining real-time warning information, it is necessary to analyze and determine whether the target area will be affected by heavy rain in a certain period of time in the future.

[0049] First, check whether the warning range in the warning information covers the target area. If it does not intersect at all, it can be judged that there will be no heavy rain in the future. If it is partially or completely covered, further confirmation is required. Secondly, determine the warning time limit, that is, the time interval between the expected start and end of the heavy rain. Check whether the interval overlaps with the future time that needs to be judged. Finally, confirm the warning level. The higher the level, the greater the danger. Generally, a yellow warning indicates that heavy rain may occur.

[0050] Taking the above factors into consideration, it can be determined whether the target area is likely to experience heavy rain in the set future time. If it is determined that there is a risk of heavy rain, flood prevention preparations need to be initiated. If there is a heavy rain, the optimal scheduling strategy combination is sent to the display end in advance at a preset time. When it is confirmed that the target area is at risk of heavy rain in the future, it is necessary to send the predetermined optimal scheduling strategy to the display end in advance to start flood prevention preparations. The display end here can be the monitoring end of the flood control system, or the terminal of the relevant operator. It is used to display the optimal strategy combination and send operating instructions to the control terminal to control the working status of infrastructure such as gates and pumping stations. The preset time refers to a certain time earlier than the expected time of heavy rain, which should be determined according to different situations. If the heavy rain is expected to have a very large rainfall, preparations should be started as early as possible. If the degree of heavy rain is average, the preset time can be shorter.

[0051] Sending the optimal strategy in advance allows operators to prepare for flood prevention work early and lock in specific operation plans. Completing system scheduling before the arrival of heavy rain can prevent disasters before they happen and minimize the impact of flood disasters that may be caused by heavy rain.

[0052] S107: Send the optimal scheduling strategy combination to a display terminal for display.

[0053] The elements involved in the determined optimal scheduling strategy, such as the control parameters of gates and pumps, execution time, etc., are encapsulated into a data packet with a unified format to facilitate network transmission and parsing. Through the application layer protocol, the encapsulated scheduling strategy data packet is sent to the display end of the flood control scheduling system, and the communication mechanism that ensures the correctness of the data is used to send instructions. After the display end obtains the data packet, it parses and extracts the specific content of the scheduling strategy, restores it to readable information, and displays the parsed scheduling strategy in a visual way on the dispatcher's human-computer interaction interface, and intuitively presents the specific scheduling content of the strategy through charts, text, etc.

[0054] In the embodiment of the present application, considering that the optimal scheduling strategies under different initial conditions may be different, a flood change prediction model based on a long short-term memory neural network is established, which can determine the corresponding optimal scheduling strategy combination for different initial rainfall conditions. This method effectively solves the problem that it is difficult to adapt to various complex situations by relying solely on artificial experience to determine the strategy, and thus realizes the scientific and accurate recommendation of the optimal flood control scheduling plan.

[0055] The following is a more detailed description of the process of the method provided by this implementation. Figure 2 , is another flow chart of the method for optimizing the scheduling of flood control and drainage by a gate and pump group in an embodiment of the present application.

[0056] S201, dividing the target area into a plurality of sub-areas according to the vegetation coverage rate in the target area; First, we need to obtain the distribution of vegetation coverage in the target area. We can analyze it through remote sensing images, and use image segmentation, classification and other technologies to determine the range of different types of vegetation. Common remote sensing images include aerospace remote sensing images and satellite remote sensing images, which have different resolutions and need to be selected according to actual needs. In addition, if there is a vegetation survey report in the target area, it can also be used as a reference.

[0057] After obtaining vegetation coverage information, it is necessary to determine the coverage rate of different vegetation types. The ratio of each type of vegetation area to the area of ​​the entire target area can be counted to obtain coverage data. Typical vegetation types can be divided into tree vegetation coverage, shrub vegetation coverage, grassland vegetation coverage, etc. Areas with higher coverage rates are divided into sub-areas. If the coverage rates of all areas are low, sub-areas can be divided according to the distribution range. Further, weights can be set according to the coverage rates of different vegetation, and areas with larger weights are used as sub-areas. It is also possible to determine the importance of different vegetation types in the description of the vegetation survey report on their impact on water volume, and set weights accordingly. Areas with larger weights can be subdivided into multiple sub-areas.

[0058] The above method can be used to divide the target area into multiple sub-areas with different vegetation coverage, which provides a basis for further considering the impact of vegetation on water volume. If the vegetation distribution in the target area is complex, the number of sub-areas can be appropriately reduced to reduce complexity.

[0059] S202, obtaining the rainwater interception and infiltration volume of each sub-area, and optimizing the flood change prediction model based on the historical flood data and gate pump operation conditions corresponding to each sub-area; For each sub-area divided in step S201, its rainwater interception and infiltration need to be estimated. Existing vegetation hydrological data can be queried. There are related studies on the theoretical or empirical values ​​of rainwater interception and infiltration of different vegetation. The interception and infiltration of specific vegetation in the target area can also be obtained through sample plot surveys, simulation tests, etc.

[0060] In addition, if there is water accumulation data in the sub-area, the actual interception of rainwater can be estimated by combining the rainfall during the flood process. If there is groundwater level monitoring, the changes under different rainfall conditions can be observed and the infiltration can be estimated. Based on historical flood and sluice pump operation data, the changes in water levels under different rainfall conditions can be analyzed, and the interception and infiltration can also be estimated.

[0061] After obtaining the rainwater interception and infiltration of each sub-region, introduce them into the flood change prediction model, adjust the model input to rainfall minus interception and infiltration, and optimize the model. You can also use interception and infiltration as new model input variables. Training the optimized model can improve the accuracy of the prediction results. After the model is optimized, verify the model effect again with historical data, observe the deviation between the model results and the actual flood situation, and if the deviation is large, adjust the model structure and parameters and retrain until the trained model can better predict the flood changes in the area.

[0062] By considering the impact of rainwater interception and infiltration of different vegetation in the region on flooding, the accuracy of flood change prediction can be improved, providing support for the subsequent determination of reasonable scheduling strategies.

[0063] S203, obtaining a plurality of water quality pollution index data according to a water quality monitoring device pre-installed in the target area; In order to evaluate and monitor the water environment quality of the target area, it is necessary to set up automatic water quality monitoring equipment or sampling points in rivers, lakes and other water bodies in the area. The monitoring equipment can use multi-parameter online water quality monitoring equipment, which can simultaneously monitor multiple indicators such as conductivity, pH, dissolved oxygen, chemical oxygen demand, ammonia nitrogen, total phosphorus, etc. The equipment records water quality parameters in real time through sensors and transmits them to the server. Sampling points need to send people to take water for testing regularly.

[0064] It is necessary to determine the location of monitoring equipment or sampling points, with a focus on being located near major water systems and covering different functional areas. Select locations based on the environmental functional zoning of the target area, such as residential areas, industrial areas, commercial areas, nature reserves, etc. After the monitoring begins, the server or humans can collect and summarize the monitoring data to form a water quality evaluation report. The report mainly includes conventional water pollution indicators such as ammonia nitrogen, total phosphorus, total nitrogen, heavy metals (such as copper, zinc, lead, cadmium, etc.), petroleum, etc. These indicators can reflect the impact of different types of pollutants on water quality.

[0065] By obtaining the above-mentioned pollution index data of the water bodies in the target area regularly or in real time, the regional water quality and the main types of pollutants can be judged, providing support for the subsequent determination of the optimal strategy for flood control and drainage.

[0066] S204, determining the polluted area according to the plurality of water pollution index data; After obtaining pollution index data sets from multiple water quality monitoring points in the target area, it is necessary to determine the actual pollution area based on these data sets. You can first conduct spatial analysis and use geographic information system software to draw a distribution map of pollution indexes at different monitoring points. By observing their spatial distribution patterns, you can preliminarily determine the scope of the river sections or waters with heavy pollution. These ranges can be divided into possible pollution areas.

[0067] Further statistical analysis can be performed on the monitoring data set to calculate the average value, maximum value, etc. of each indicator, and compare them with the corresponding evaluation criteria or thresholds to screen out the monitoring points that exceed the standard. These points that exceed the standard are more likely to belong to the polluted area. For the range of possible polluted areas, the arithmetic mean of the pollution indicators of all monitoring points in the area can be calculated. If the average value exceeds the standard, it will be determined as a polluted area. If it does not exceed the standard, it is necessary to check the value of a single monitoring point. As long as there is a monitoring point that exceeds the standard, it can also be determined as a polluted area.

[0068] In addition, different functional zones such as industrial areas can be directly identified as polluted areas if the corresponding monitoring points have high data values, and the monitoring point data of sensitive areas such as residential areas and drinking water sources should be focused on. By comprehensively utilizing spatial analysis, statistical analysis and other means, based on water quality monitoring data sets, the main polluted areas in the target area can be identified, providing a basis for considering water quality impacts when formulating subsequent flood control and drainage strategies.

[0069] S205, setting a regional weight according to the water quality pollution index data in the polluted area and a set threshold; According to the polluted area determined in step S204, it is necessary to access the data of the water quality monitoring points within the area. According to different water quality pollution indicators, corresponding evaluation thresholds are pre-set.

[0070] The following method can be used to set regional weights: obtain all pollution index monitoring data of each monitoring point in the polluted area, and for each pollution index, calculate the arithmetic mean of the data of all monitoring points in the polluted area. Compare the average value with the corresponding index threshold. If the average value exceeds the threshold, the weight of each monitoring point that exceeds the index threshold is increased by 1, and all index weights of each monitoring point are summarized to obtain the comprehensive weight of each monitoring point. The comprehensive weights of all monitoring points in the polluted area are averaged to obtain the regional weight of the polluted area. Other evaluation functions can also be used, such as eradicating the sum of squares of the degree of excess of different indicators, to comprehensively determine the weight. The weight can be based on a 1-10 rating system, and the regional weight represents the comprehensive pollution severity of the polluted area.

[0071] By setting regional weights, the relative importance of each polluted area can be determined from the perspective of pollution degree, which can provide a reference for the subsequent formulation of flood control and drainage strategies so that the strategies take into account the pollution control needs of different regions.

[0072] S206, adjusting and optimizing the flood change prediction model according to the weight of the region.

[0073] After determining the weight of each polluted area, the optimized flood change prediction model can be further adjusted according to the weight. A mapping table that associates each sub-area with the corresponding polluted area can be constructed. When performing flood simulation prediction, for each sub-area in the prediction result, check the weight of the mapped polluted area.

[0074] If the weight is high, the predicted flood volume of the sub-region is reduced in the model results. This can make the model biased towards choosing flood control strategies that will not spread floods to heavily polluted areas. It is also possible to establish a mathematical model of the sub-region and the pollution weight and introduce it into the model as an additional constraint or objective function. For example, minimize the amount of flood discharge overflowing into highly polluted areas. The adjusted model will consider the potential impact on polluted areas when simulating different flood control and drainage strategies. This can avoid choosing strategies that spread pollution and make flood control and drainage more environmentally friendly. The adjusted model has been used for verification many times, and the results can meet the needs of flood control while avoiding flood discharge to heavily polluted areas, improving the scientific nature of the strategy.

[0075] In the embodiment of the present application, since the vegetation coverage and water pollution conditions in different sub-areas in the target area are taken into consideration, the target area is divided into sub-areas according to the vegetation coverage rate, the rainwater interception and infiltration amount of each sub-area are obtained, and the polluted areas and their weights are determined based on water quality monitoring. Therefore, the vegetation and water quality influences are effectively introduced into the flood change prediction model, the model is optimized, and the prediction deviation problem caused by the original model not considering these factors is solved, thereby improving the accuracy of flood change prediction and the scientific nature of flood control scheduling strategies.

[0076] The following describes the scheduling server in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , which is a schematic diagram of a physical device structure of a scheduling server in an embodiment of the present application.

[0077] It should be noted that Figure 3 The structure of the scheduling server shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0078] like Figure 3 As shown, the scheduling server includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (RAM) 303, such as executing the method described in the above embodiment. In RAM 303, various programs and data required for system operation are also stored. CPU 301, ROM 302 and RAM 303 are connected to each other through bus 304. Input / output (I / O) interface 305 is also connected to bus 304.

[0079] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a button switch, etc.; an output section 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0080] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 309, and / or installed from a removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0081] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0083] Specifically, the dispatch server of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the optimized dispatching method for flood control and drainage of a gate and pump group provided in the above embodiment is implemented.

[0084] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the dispatch server described in the above embodiment; or may exist independently without being assembled into the dispatch server. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the dispatch server, the dispatch server implements the flood control and drainage optimization dispatch method of the gate pump group provided in the above embodiment.

[0085] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0086] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0087] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. A flood control and drainage optimization scheduling method for a gate pump group, applied to a scheduling server, characterized in that: The method comprises: Obtain historical flood data of the target area at multiple historical time points and corresponding historical gate pump operation data, wherein the historical gate pump operation data includes gate opening and pump start / stop status; Inputting the historical flood data and the historical gate pump operation data into a long short-term memory neural network for processing, and constructing a flood change prediction model, wherein the flood change prediction model is used for flood changes under different scheduling strategy combinations, wherein the scheduling strategy combination includes different combinations of the gate opening and the start and stop states of the pump; After setting different initial conditions, multiple simulation calculations are performed through the flood change prediction model to determine the flood change prediction under multiple scheduling strategy combinations, wherein the initial conditions include different rainfall conditions; Determine the flood control and drainage effect of the target area under the scheduling strategy combination according to the flood change prediction, and the flood control and drainage effect at least includes the water level control condition; After determining the optimal scheduling strategy combination according to the flood control and drainage effect, using the optimal scheduling strategy combination and the corresponding initial conditions as training samples to continue training the long short-term memory neural network; Receiving real-time meteorological and hydrological data of the target area, extracting actual initial conditions, and inputting the actual initial conditions into the long short-term memory neural network to determine the actual optimal scheduling strategy combination; The optimal scheduling strategy combination is sent to the display terminal for display.

2. The method according to claim 1, characterized in that After receiving the real-time meteorological and hydrological data of the target area and extracting the actual initial conditions, the actual initial conditions are input into the long short-term memory neural network to determine the actual optimal scheduling strategy combination, and further comprising: Receiving weather warnings from a meteorological terminal in the target area; Determine whether there is heavy rain within the set time based on the weather warning; If there is a rainstorm, the optimal scheduling strategy combination is sent to the display terminal in advance at a preset time; Continuously track changes in weather warning information and dynamically combine the optimal scheduling strategies.

3. The method according to claim 1, characterized in that After the step of inputting the historical flood data and the historical sluice pump operation data into the long short-term memory neural network for processing and constructing the flood change prediction model, the method further includes: Dividing the target area into a plurality of sub-areas according to the vegetation coverage rate in the target area; The rainwater interception and infiltration volume of each sub-area are obtained, and the flood change prediction model is optimized by combining the historical flood data and gate pump operation conditions corresponding to each sub-area.

4. The method according to claim 3, characterized in that After obtaining the rainwater interception and infiltration of each sub-area, and optimizing the flood change prediction model in combination with the historical flood data and the operation status of the sluice pump corresponding to each sub-area, the method further includes: Acquiring a plurality of water quality pollution index data according to water quality monitoring equipment pre-installed in the target area; Determine the polluted area according to the plurality of water pollution index data; Setting a regional weight according to the water quality pollution index data in the polluted area and a set threshold; The flood change prediction model is optimized according to the regional weight adjustment.

5. The method according to claim 1, characterized in that After the step of determining the flood control and drainage effect of the target area under the scheduling strategy combination according to the flood change prediction situation, the method further includes: Determine the energy efficiency of the pump under each of the scheduling strategy combinations, wherein the energy efficiency is determined by the ratio of the displacement of the pump to the energy consumption; The flood control and drainage effect is adjusted according to the energy utilization efficiency.

6. The method according to claim 5, characterized in that After the step of adjusting the flood control and drainage effect according to the energy utilization efficiency, the method further includes: Determine the total time from the start of execution of each scheduling strategy combination to achieving the expected effect of flood control and drainage under different initial conditions; The flood control and drainage effect is adjusted again according to the total time.

7. The method according to claim 1, characterized in that After determining the optimal scheduling strategy combination according to the flood control and drainage effect, taking the optimal scheduling strategy combination and the corresponding initial conditions as training samples, and continuing to train the long short-term memory neural network, the method further includes: After determining the impact of different scheduling strategy combinations on the upstream and downstream areas, evaluating the different scheduling strategy combinations according to preset upstream and downstream impact evaluation indicators; If the scheduling strategy combination causes multiple indicators of the upstream and downstream areas to exceed the corresponding preset thresholds of the indicators, the corresponding scheduling strategy combination is marked as an adverse impact strategy; When subsequently training the long short-term memory neural network, the weight of the adverse impact strategy in the training sample is reduced.

8. A server, characterized in that: The server includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the server to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a server, the server is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is run on a server, the server is caused to execute the method according to any one of claims 1 to 7.

Images