Self-adaptive integrated pump station dynamic optimization control method and system

Abstract

The invention discloses a self-adaptive integrated pump station dynamic optimization control method and system, relates to the technical field of water pump station intelligent control, and solves the problems that a control strategy adopting a preset rule in the prior art is difficult to adapt to the real-time working condition change of a pump station operation environment, and the working efficiency is low. The energy efficiency of the pump station control method is not high; and the maintenance cost is increased. Predicted water inlet flow is generated according to pump station data and environment data; generating a device state coefficient based on the device parameter; generating a working condition state label based on the predicted water inlet flow; based on the equipment state coefficient and the working condition state label, constructing a multi-objective optimization function and generating an alarm signal; obtaining a control instruction based on the multi-objective optimization function; according to the method, related equipment is dynamically controlled based on a control instruction, pump station water inlet data in a period of time in the future is predicted, and a target optimization function is constructed in the directions of energy consumption and cost, so that energy consumption and cost minimization is achieved, and the efficiency of the pump station control method is improved.

Description

Technical Field

[0001] The present application belongs to the field of intelligent control technology for water pumping stations, and specifically is a method and system for dynamic optimization control of an adaptive integrated pumping station. Background Art

[0002] Pump stations are key facilities in urban drainage, flood control, sewage treatment, and water supply systems. Their operational efficiency and stability directly impact the smooth functioning of the entire system. As critical hydraulic regulation and transmission nodes, pump stations play a vital role in ensuring the normal operation of cities and improving the efficiency of water resource utilization. With the acceleration of urbanization, the continuous upgrading and optimization of pump station control technology has become a key direction for improving urban water management.

[0003] Traditional pump station control systems mostly use control strategies based on preset rules, which are difficult to adapt to real-time operating condition changes in the pump station operating environment, resulting in low energy efficiency of pump station control methods and increased maintenance costs; therefore, further improvements are still needed for pump station control methods. Summary of the Invention

[0004] The present application aims to solve at least one of the technical problems existing in the prior art; to this end, the present application proposes an adaptive integrated pump station dynamic optimization control method and system to solve the technical problem that the prior art adopts a control strategy with preset rules, which is difficult to adapt to the real-time working condition changes of the pump station operating environment, resulting in low energy efficiency of the pump station control method and increased maintenance costs.

[0005] To achieve the above objectives, the first aspect of the present application provides an adaptive integrated pump station dynamic optimization control method, comprising:

[0006] Obtain equipment parameters, environmental data, and pump station data;

[0007] Generates a predicted inflow flow based on pump station data and environmental data; the predicted inflow flow refers to the amount of water entering the pump station over a period of time in the future;

[0008] Generate a device status coefficient based on device parameters; the device status coefficient is used to quantify the usage status of the device;

[0009] Generate an operating status label based on the predicted water inlet flow rate; the operating status label refers to a label measuring the predicted flow rate;

[0010] Construct a multi-objective optimization function based on equipment status coefficients and working condition labels and generate alarm signals;

[0011] Obtaining control instructions based on a multi-objective optimization function;

[0012] Dynamically control related equipment based on control instructions.

[0013] Through the above steps, this application predicts the water inflow situation of the pumping station in the future, and constructs an optimization function based on the prediction results with the goal of minimizing energy consumption and cost, thereby effectively optimizing the pumping station control method and improving its operating efficiency.

[0014] Furthermore, generating a predicted water inflow flow rate based on the pump station data and the environmental data includes:

[0015] The environmental data includes environmental parameters and weather forecasts;

[0016] Extract the inlet flow, time characteristics, environmental parameters and equipment parameters corresponding to several recorded time points within the sliding time window of the pump station data;

[0017] Integrate the water inlet flow, time characteristics, environmental parameters and equipment parameters corresponding to the same recording time point into recording time data;

[0018] Integrate the recorded time data and weather forecast data of several recorded time points into a water inflow prediction sequence;

[0019] The water inlet flow prediction sequence is input into the flow prediction model to obtain the predicted water inlet flow; the flow prediction model is constructed using the LSTM model.

[0020] Furthermore, the traffic prediction model is constructed using an LSTM model, including:

[0021] Obtain several historical inlet flow prediction sequences and their corresponding inlet flow rates;

[0022] Divide several historical inflow flow prediction sequences and their corresponding inflow flows into training data, validation data, and test data; and perform data preprocessing on the training data, validation data, and test data to obtain a training set, a validation set, and a test set;

[0023] Select the LSTM model as the base model;

[0024] Train the basic model using the training set, and adjust the learning rate and other hyperparameters on the validation set to obtain the pre-trained model;

[0025] By verifying the pre-trained model on the test set, we finally obtained a flow prediction model whose input is the inlet flow prediction sequence and output is the predicted inlet flow.

[0026] Furthermore, generating the device state coefficient based on the device parameters includes:

[0027] The equipment parameters include the equipment ID and its corresponding equipment type, the remaining life ratio SSB, the current efficiency ratio DXB, the degradation trend coefficient LQX, and the maintenance coefficient WX. The remaining life ratio quantifies the remaining service life of the equipment ID. The specific calculation formula is as follows: Where LYS and SS represent the cumulative operating time and total design time of the device ID, respectively; ZGC represents the number of major failures; k represents the failure impact coefficient, where k∈(0,1); the current efficiency ratio evaluates the instantaneous operating performance of the device, and is calculated as follows: Wherein, DX and EX represent the current efficiency and rated efficiency corresponding to the device ID, respectively. The degradation trend coefficient quantifies the real-time deviation trend of the device ID and is calculated as follows: Where i represents the number of the influencing factor, CPL represents the parameter deviation rate of the influencing factor, and max() represents the maximum value operation. The maintenance coefficient quantifies the maintenance status of the equipment ID and is calculated as follows: WX = 0.6 × BJW + 0.4 × PXSY, which comprehensively considers the two influencing factors of maintenance record completeness BJW and mean repair time factor PXSY.

[0028] The device status coefficient SZX corresponding to the device ID is calculated using the formula SZX = α1 × (1-SSB) + α2 × DXB + α3 × LQX + α4 × WX; where α1, α2, α3, and α4 are weight coefficients, α1, α2, α3, and α4∈(0,1), and the sum of the four weight coefficients α1, α2, α3, and α4 is 1;

[0029] Based on the device type, several weight coefficients corresponding to the current device ID are obtained from a weight recommendation table; the weight recommendation table is set by an expert based on experience.

[0030] This application considers multi-angle data from each device and comprehensively evaluates the status level corresponding to the device ID, making the status evaluation corresponding to each device ID more accurate and providing accurate data support for the subsequent generation of control strategies.

[0031] Furthermore, generating the operating status label based on the predicted water inlet flow rate includes:

[0032] Extract the water inlet flow JL corresponding to the pump station data;

[0033] By formula Calculate the flow rate change rate LBL; where YJL represents the predicted water inlet flow;

[0034] When the flow rate change rate is greater than or equal to the first change rate, the operating state label is set to a mutation label;

[0035] When the flow rate change rate is less than the first change rate and greater than or equal to the second change rate, the operating state label is set to a gradual change label;

[0036] When the flow rate change rate is less than the second change rate, setting the operating state label to a stable label; wherein the first change rate is greater than the second change rate;

[0037] The operating state labels include sudden change labels, gradual change labels and stable labels.

[0038] Furthermore, the multi-objective optimization function is constructed based on the equipment state coefficient and the working condition state label and the alarm signal is generated, including:

[0039] Construct a multi-objective optimization function; the multi-objective optimization function satisfies the following formula:

[0040] Among them, j represents the number corresponding to the optimization target, J represents the total number of optimization targets, β j It is expressed as the optimization weight coefficient corresponding to the jth optimization target; the optimization weight coefficient is generated by the equipment status coefficient and the working condition status label; YMF j It is expressed as the target optimization function corresponding to the j-th optimization goal;

[0041] When the device status coefficient is less than the status threshold, a device damage alarm signal is generated for the device ID corresponding to the device status coefficient;

[0042] When the operating status tag is a sudden change tag, a water surge alarm signal for the pump station is generated.

[0043] Furthermore, the optimization weight coefficient is generated by the equipment status coefficient and the working condition status label, including:

[0044] Extracting the operating status label, equipment status coefficient, electricity price period label, and environmental policy factor; the electricity price period label includes peak electricity price period, valley electricity price period, and flat electricity price period; the environmental policy factor refers to the quantification degree of the current environmental status;

[0045] Integrate operating status labels, equipment status coefficients, electricity price period labels, and environmental policy factors into optimization analysis data;

[0046] The optimization analysis data is input into the optimization weight generation model to obtain the optimization weights corresponding to the target optimization functions; the optimization weight generation model is constructed by an artificial intelligence model, including:

[0047] Obtaining some historical optimization analysis data and their corresponding historical optimization weights;

[0048] Divide a number of historical optimization analysis data and their corresponding historical optimization weights into training data, verification data, and test data; and perform data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set;

[0049] Select an artificial intelligence model as the base model;

[0050] Train the basic model using the training set, and adjust the learning rate and other hyperparameters on the validation set to obtain the pre-trained model;

[0051] By verifying the pre-trained model on the test set, we finally obtain an optimization weight generation model whose input is optimization analysis data and output is the optimization weights corresponding to several target optimization functions.

[0052] Furthermore, the target optimization function is constructed in the following manner, including:

[0053] Extract the water inlet flow JL, head YC and operating efficiency YX corresponding to the equipment ID in the equipment parameters;

[0054] Construct the energy consumption target optimization function NHF, which satisfies the following formula:

[0055] Where m represents the number corresponding to the device ID, ΔYS represents the device operating time, SM and SG represent the water density and gravity acceleration, respectively;

[0056] Obtaining electricity prices DJ in different time periods, where the electricity prices are obtained through an electricity price schedule;

[0057] Extract the device power SP and device operating time ΔYS corresponding to each device ID in different time periods, as well as the number of starts and stops QTC and cumulative operating time LYS corresponding to each device ID;

[0058] Construct a cost target optimization function CBF, which satisfies the following formula:

[0059]

[0060] Where m represents the number corresponding to the device ID, t represents the number of the time period, c1 and c2 represent the maintenance cost coefficients, c1 represents the cost of a single start and stop, and c2 represents the maintenance cost per unit operating time;

[0061] Construct the environmental target optimization function HJF, which satisfies the following formula:

[0062] HJF = tp × NHF; where tp represents the carbon emission factor;

[0063] The target optimization function includes an energy consumption target optimization function, a cost target optimization function and an environmental target optimization function.

[0064] Furthermore, obtaining the control instruction based on the multi-objective optimization function includes:

[0065] Solving the multi-objective optimization function through a hierarchical optimization strategy to obtain a decision variable; the decision variable refers to the best result that satisfies the multi-objective optimization function under the current state;

[0066] Parse the decision variables into control set values ​​for each device according to predefined meanings;

[0067] Map control setting values ​​to control instructions for each device ID.

[0068] Another aspect of the present invention provides an adaptive integrated pump station dynamic optimization control system, comprising: a data acquisition module, a data analysis module, a control module and an early warning module; the data acquisition module is connected to the data analysis module; the data analysis module is connected to the early warning module; the data analysis module is connected to the control module;

[0069] The data acquisition module is used to acquire equipment parameters, environmental data and pump station data through data acquisition equipment;

[0070] The data analysis module generates a predicted water inflow based on pump station data and environmental data; generates a device status coefficient based on device parameters; generates an operating status label based on the predicted water inflow; constructs a multi-objective optimization function based on the device status coefficient and the operating status label and generates an alarm signal; and obtains a control instruction based on the multi-objective optimization function.

[0071] The control module dynamically controls the relevant equipment based on the control instructions;

[0072] The early warning module issues a prompt according to the alarm signal.

[0073] Compared with the prior art, the present invention has the following advantages:

[0074] 1. This application generates a predicted water inlet flow rate based on pump station data and environmental data; generates an equipment status coefficient based on equipment parameters; generates an operating condition label based on the predicted water inlet flow rate; constructs a multi-objective optimization function based on the equipment status coefficient and the operating condition label and generates an alarm signal; obtains control instructions based on the multi-objective optimization function; dynamically controls related equipment based on the control instructions, predicts the pump station water inlet data in the future, and constructs a target optimization function in the direction of energy consumption and cost, so as to minimize energy consumption and cost and improve the efficiency of the pump station control method.

[0075] 2. This application considers multi-angle data from each device and comprehensively evaluates the status level corresponding to the device ID, making the status evaluation corresponding to each device ID more accurate and providing accurate data support for the subsequent generation of control strategies.

[0076] 3. This application improves the accuracy of the multi-objective optimization function by constructing a multi-objective optimization function and dynamically adjusting the corresponding optimization weights for each objective optimization function through a pre-trained optimization weight generation model; at the same time, the multi-objective optimization function considers energy consumption, cost and environment from three aspects, which can maximize the efficiency of the pump station control method to the greatest extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0078] Figure 1 This is a flow chart of a dynamic optimization control method for an adaptive integrated pump station of the present application;

[0079] Figure 2 This is a schematic diagram of the principle of an adaptive integrated pump station dynamic optimization control system of this application. DETAILED DESCRIPTION

[0080] The following will clearly and completely describe the technical solutions of this application in conjunction with the embodiments. Obviously, the embodiments described are only a part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0081] See also Figure 1 The first embodiment of the present application provides an adaptive integrated pump station dynamic optimization control method, including:

[0082] Obtain equipment parameters, environmental data, and pump station data;

[0083] Generates a predicted inflow flow based on pump station data and environmental data; the predicted inflow flow refers to the amount of water entering the pump station over a period of time in the future;

[0084] Generate a device status coefficient based on device parameters; the device status coefficient is used to quantify the usage status of the device;

[0085] Generate an operating status label based on the predicted water inlet flow rate; the operating status label refers to a label measuring the predicted flow rate;

[0086] Construct a multi-objective optimization function based on equipment status coefficients and working condition labels and generate alarm signals;

[0087] Obtaining control instructions based on a multi-objective optimization function;

[0088] Dynamically control related equipment based on control instructions.

[0089] In this embodiment, the predicted water inflow rate is generated based on the pump station data and the environmental data, including:

[0090] Environmental data includes environmental parameters and weather forecasts. Weather forecasts refer to the predictions of environmental data for a period of time in the future. Environmental parameters include temperature, humidity, and wind speed.

[0091] Extract the water inflow, time characteristics, environmental parameters, and equipment parameters corresponding to several recorded time points within a sliding time window in the pump station data. The size of the sliding time window is set based on experience. In this embodiment, the sliding time window is set to 72 hours. In other embodiments, it can be set to 24 hours or even 96 hours. The time characteristics include seasonal characteristics and weekday characteristics.

[0092] Integrate the water inlet flow, time characteristics, environmental parameters and equipment parameters corresponding to the same recording time point into recording time data;

[0093] Integrate the recorded time data and weather forecast data of several recorded time points into a water inflow prediction sequence;

[0094] The inlet flow prediction sequence is input into the flow prediction model to obtain the predicted inlet flow; the flow prediction model is constructed through the LSTM model to estimate the inlet flow of the pump station in the future.

[0095] The traffic prediction model in this embodiment is constructed using the LSTM model, including:

[0096] Obtain several historical inlet flow prediction sequences and their corresponding inlet flow rates;

[0097] Several historical inflow flow prediction sequences and their corresponding inflow flows are divided into training data, validation data, and test data; data preprocessing is performed on the training data, validation data, and test data to obtain training sets, validation sets, and test sets; the ratio of the training set, test set, and validation set is 7:2:1;

[0098] Select the LSTM model as the base model;

[0099] Train the basic model using the training set, and adjust the learning rate and other hyperparameters on the validation set to obtain the pre-trained model;

[0100] By verifying the pre-trained model on the test set, we finally obtained a flow prediction model whose input is the inlet flow prediction sequence and output is the predicted inlet flow.

[0101] This embodiment uses a pre-trained flow prediction model to predict the water inlet flow of the pump station in the future, comprehensively considers multiple influencing factors, significantly improves the accuracy of the water inlet flow prediction, and provides reliable data support for the generation of subsequent control strategies.

[0102] In this embodiment, the device state coefficient is generated based on the device parameters, including:

[0103] Equipment parameters include the device ID and its corresponding device type, remaining life ratio (SSB), current efficiency ratio (DXB), degradation trend coefficient (LQX), and maintenance coefficient (WX). The remaining life ratio quantifies the remaining service life of the device ID, taking into account the proportion of years the device has been in use and the reduction in service life due to serious failures. The specific calculation formula is as follows: Where LYS and SS represent the cumulative operating time and designed total operating time of the device ID, respectively; ZGC represents the number of major failures; k represents the failure impact coefficient, k∈(0,1). The specific value is set based on experience. In this embodiment, k is set to 0.1; the current efficiency ratio evaluates the real-time operating performance of the device, and the calculation formula is as follows: Where DX and EX represent the current efficiency and rated efficiency corresponding to the device ID, respectively. The degradation trend coefficient quantifies the real-time deviation trend of the device ID and is calculated as follows: Where i represents the number corresponding to the influencing factor, CPL represents the parameter offset rate corresponding to the influencing factor, and max() represents the maximum value operation. Usually, 3-5 influencing factors are selected for consideration. In this embodiment, vibration offset rate, temperature offset rate, and current harmonic offset rate are considered. The maintenance coefficient is the maintenance status of the quantified equipment ID. The calculation formula is as follows: WX = 0.6 × BJW + 0.4 × PXSY, which comprehensively considers the two influencing factors of maintenance record completeness BJW and mean repair time factor PXSY.

[0104] The device status coefficient SZX corresponding to the device ID is calculated using the formula SZX = α1 × (1-SSB) + α2 × DXB + α3 × LQX + α4 × WX; where α1, α2, α3, and α4 are weight coefficients, α1, α2, α3, and α4∈(0,1), and the sum of the four weight coefficients α1, α2, α3, and α4 is 1;

[0105] Based on the device type, obtain the weight coefficients corresponding to the current device ID from the weight recommendation table;

[0106] The weight recommendation table is set by experts based on experience. The weight recommendation table of this embodiment is as follows:

[0107]

[0108] In this embodiment, if the equipment type is a core main pump, when calculating its corresponding equipment status coefficient, the selected weight coefficients are 0.4, 0.35, 0.15 and 0.1.

[0109] This embodiment comprehensively analyzes the multi-dimensional data from each device and evaluates the status level corresponding to each device ID, thereby improving the accuracy of status assessment and providing more reliable data support for the generation of subsequent control strategies.

[0110] In this embodiment, the operating status label is generated based on the predicted water inflow rate, including:

[0111] Extract the water inlet flow JL corresponding to the pump station data;

[0112] By formula Calculate the flow rate change rate LBL; where YJL represents the predicted water inlet flow;

[0113] When the flow rate change rate is greater than or equal to the first change rate, the operating state label is set to a mutation label;

[0114] When the flow rate change rate is less than the first change rate and greater than or equal to the second change rate, the operating state label is set to a gradual change label;

[0115] When the flow rate change rate is less than the second change rate, the operating state label is set to a stable label; wherein the first change rate is greater than the second change rate, and the specific values ​​are set based on experience. In this embodiment, the first change rate and the second change rate are set to 0.5 and 0.2 respectively; in other embodiments, the first change rate and the second change rate can be set to 0.6 and 0.3 respectively;

[0116] The working condition status labels include sudden change labels, gradual change labels and stable labels.

[0117] In this embodiment, a multi-objective optimization function is constructed based on the equipment state coefficient and the operating condition label and an alarm signal is generated, including:

[0118] Construct a multi-objective optimization function; the multi-objective optimization function satisfies the following formula:

[0119] Among them, j represents the number corresponding to the optimization target, J represents the total number of optimization targets, β jIt is expressed as the optimization weight coefficient corresponding to the jth optimization target; the optimization weight coefficient is generated by the equipment status coefficient and the working condition status label; YMF j It is represented as the target optimization function corresponding to the jth optimization goal. In this embodiment, there are three optimization goals, namely, energy efficiency maximization, cost minimization, and environmental impact minimization, so J is 3. Since energy efficiency maximization is equivalent to energy consumption minimization, energy consumption minimization is used instead of energy efficiency maximization when constructing the multi-objective function.

[0120] When the device status coefficient is less than the status threshold, a device damage alarm signal is generated for the device ID corresponding to the device status coefficient. The status threshold is set based on experience. In this embodiment, the status threshold is set to 0.5.

[0121] When the operating status tag is a sudden change tag, a water surge alarm signal for the pump station is generated.

[0122] The optimization weight coefficient in this embodiment is generated by the equipment status coefficient and the working condition status label, including:

[0123] Extract operating status labels, equipment status coefficients, electricity price period labels, and environmental policy factors; electricity price period labels include peak electricity price periods, valley electricity price periods, and flat electricity price periods; environmental policy factors refer to the quantification of the current environmental status. If local environmental protection requirements are strict, such as a heavy pollution weather warning, the optimization weight corresponding to the environmental objective optimization function will be increased;

[0124] Integrate operating status labels, equipment status coefficients, electricity price period labels, and environmental policy factors into optimization analysis data;

[0125] The optimization analysis data is input into the optimization weight generation model to obtain the optimization weights corresponding to several target optimization functions; the optimization weight generation model is constructed through an artificial intelligence model, including:

[0126] Obtaining some historical optimization analysis data and their corresponding historical optimization weights;

[0127] Divide a number of historical optimization analysis data and their corresponding historical optimization weights into training data, validation data, and test data; perform data preprocessing on the training data, validation data, and test data to obtain a training set, a validation set, and a test set; the ratio of the training set, the test set, and the validation set is 7:2:1;

[0128] Select an artificial intelligence model as the basic model; artificial intelligence models include BP models, etc.

[0129] Train the basic model using the training set, and adjust the learning rate and other hyperparameters on the validation set to obtain the pre-trained model;

[0130] By verifying the pre-trained model on the test set, we finally obtain an optimization weight generation model whose input is optimization analysis data and output is the optimization weights corresponding to several target optimization functions.

[0131] The target optimization function in this embodiment is constructed in the following manner, including:

[0132] Extract the water inlet flow JL, head YC and operating efficiency YX corresponding to the equipment ID in the equipment parameters;

[0133] Construct the energy consumption target optimization function NHF, which satisfies the following formula:

[0134] Where m represents the number corresponding to the device ID, ΔYS represents the device operating time, SM and SG represent the water density and gravity acceleration, respectively;

[0135] Get the electricity price DJ for different time periods. The electricity price is obtained through the electricity price schedule; the electricity price schedule is:

[0136]

[0137] Extract the device power SP and device operating time ΔYS corresponding to each device ID in different time periods, as well as the number of starts and stops QTC and cumulative operating time LYS corresponding to each device ID;

[0138] Construct a cost target optimization function CBF, which satisfies the following formula:

[0139]

[0140] Where m represents the number corresponding to the device ID, t represents the number of the time period, c1 and c2 represent the maintenance cost coefficients, c1 is used to represent the cost of a single start and stop, and c2 is used to represent the maintenance cost per unit operating time. The specific values ​​are set based on experience. In this embodiment, c1 is set to 5 yuan / time and c2 is set to 1 yuan / hour.

[0141] Construct the environmental target optimization function HJF, which satisfies the following formula:

[0142] HJF = tp × NHF; where tp represents the carbon emission factor. The specific value is set based on experience. In this embodiment, tp is set to 0.583 kgCO2 / kWh;

[0143] The target optimization function includes energy consumption target optimization function, cost target optimization function and environmental target optimization function;

[0144] The objective optimization function has the following constraints:

[0145] The total water inlet flow must be equal to the predicted water inlet flow, the water inlet flow and head of the equipment must be within the operating range, the equipment speed must be within the allowable range, and the equipment start and stop must meet the minimum start and stop time to avoid frequent start and stop behavior; when the equipment status coefficient is lower than the status threshold, the equipment will not start.

[0146] This embodiment improves the accuracy and adaptability of the multi-objective optimization function by constructing a multi-objective optimization function and using a pre-trained weight generation model to dynamically adjust the optimization weight corresponding to each objective. The optimization function comprehensively considers the three dimensions of energy consumption, cost, and environment, while achieving efficient operation of the pump station control method and maximizing the overall system performance.

[0147] In this embodiment, the control instructions are obtained based on the multi-objective optimization function, including:

[0148] The decision variables are obtained by solving the multi-objective optimization function through the hierarchical optimization strategy; the decision variables refer to the best results that satisfy the multi-objective optimization function under the current state;

[0149] Parse the decision variables into control set values ​​for each device according to predefined meanings;

[0150] Map control setting values ​​to control instructions for each device ID.

[0151] See also Figure 2 Another aspect of the present application provides an adaptive integrated pump station dynamic optimization control system, including: a data acquisition module, a data analysis module, a control module and an early warning module; the data acquisition module is connected to the data analysis module; the data analysis module is connected to the early warning module; the data analysis module is connected to the control module;

[0152] Data acquisition module: acquires equipment parameters, environmental data and pump station data through data acquisition equipment; the data acquisition equipment includes several sensors, etc.

[0153] Data analysis module: Generates predicted water inflow based on pump station data and environmental data; generates equipment status coefficients based on equipment parameters; generates operating status labels based on predicted water inflow; constructs a multi-objective optimization function based on the equipment status coefficients and operating status labels and generates alarm signals; and obtains control instructions based on the multi-objective optimization function;

[0154] Control module: dynamically controls related equipment based on control instructions;

[0155] Early warning module: issues prompts based on alarm signals; alarm signals include equipment damage alarm signals and pump station water volume surge alarm signals, etc.

[0156] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0157] The working principle of this application is: by obtaining equipment parameters, environmental data and pump station data; generating a predicted water inlet flow rate according to the pump station data and environmental data; generating an equipment status coefficient based on the equipment parameters; generating an operating condition status label based on the predicted water inlet flow rate; constructing a multi-objective optimization function based on the equipment status coefficient and the operating condition status label and generating an alarm signal; obtaining control instructions based on the multi-objective optimization function; dynamically controlling related equipment based on the control instructions to predict the pump station water inlet data in the future, and constructing a target optimization function in the direction of energy consumption and cost, so as to minimize energy consumption and cost, improve the efficiency of the pump station control method, and avoid the control strategy of the existing technology using preset rules, which is difficult to adapt to the real-time operating condition changes of the pump station operating environment, resulting in low energy efficiency of the pump station control method and increased maintenance costs.

[0158] The above embodiments are only used to illustrate the technical method of the present application and are not intended to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present application.

Claims

1. An adaptive integrated pump station dynamic optimization control method, characterized in that: include: Obtain equipment parameters, environmental data, and pump station data; Generates predicted inflow flow based on pump station data and environmental data; The predicted inflow refers to the amount of water entering the pumping station in the future; Generate a device status coefficient based on device parameters; the device status coefficient is used to quantify the usage status of the device; Generate operating status labels based on predicted inlet flow; The working condition label refers to the measurement label of the flow rate for prediction; Construct a multi-objective optimization function based on equipment status coefficients and working condition labels and generate alarm signals; Obtaining control instructions based on a multi-objective optimization function; Dynamically control related equipment based on control instructions.

2. The adaptive integrated pump station dynamic optimization control method according to claim 1, characterized in that: The method of generating a predicted water inflow flow rate based on the pump station data and the environmental data includes: The environmental data includes environmental parameters and weather forecasts; Extract the inlet flow, time characteristics, environmental parameters and equipment parameters corresponding to several recorded time points within the sliding time window of the pump station data; Integrate the water inlet flow, time characteristics, environmental parameters and equipment parameters corresponding to the same recording time point into recording time data; Integrate the recorded time data and weather forecast data of several recorded time points into a water inflow prediction sequence; The water inlet flow prediction sequence is input into the flow prediction model to obtain the predicted water inlet flow; the flow prediction model is constructed using the LSTM model.

3. The adaptive integrated pump station dynamic optimization control method according to claim 2, characterized in that: The traffic prediction model is constructed using the LSTM model, including: Obtain several historical inlet flow prediction sequences and their corresponding inlet flow rates; Divide several historical inflow flow prediction sequences and their corresponding inflow flows into training data, validation data, and test data; and perform data preprocessing on the training data, validation data, and test data to obtain a training set, a validation set, and a test set; Select the LSTM model as the base model; Train the basic model using the training set, and adjust the learning rate and other hyperparameters on the validation set to obtain the pre-trained model; By verifying the pre-trained model on the test set, we finally obtained a flow prediction model whose input is the inlet flow prediction sequence and output is the predicted inlet flow.

4. The adaptive integrated pump station dynamic optimization control method according to claim 1, characterized in that: Generating the device state coefficient based on the device parameters includes: The equipment parameters include the equipment ID and its corresponding equipment type, the remaining life ratio SSB, the current efficiency ratio DXB, the degradation trend coefficient LQX, and the maintenance coefficient WX; the remaining life ratio quantifies the remaining service life of the equipment ID; the current efficiency ratio evaluates the real-time operating performance of the equipment; the degradation trend coefficient quantifies the real-time deviation trend of the equipment ID; and the maintenance coefficient quantifies the maintenance status of the equipment ID. The device status coefficient SZX corresponding to the device ID is calculated using the formula SZX = α1 × (1-SSB) + α2 × DXB + α3 × LQX + α4 × WX; where α1, α2, α3, and α4 are weight coefficients, α1, α2, α3, and α4∈(0,1), and the sum of the four weight coefficients α1, α2, α3, and α4 is 1; Based on the device type, several weight coefficients corresponding to the current device ID are obtained from a weight recommendation table; the weight recommendation table is set by an expert based on experience.

5. The adaptive integrated pump station dynamic optimization control method according to claim 1, characterized in that: The generating of the operating condition label based on the predicted water inlet flow rate includes: Extract the water inlet flow JL corresponding to the pump station data; By formula Calculate the flow rate change rate LBL; where YJL represents the predicted water inlet flow; When the flow rate change rate is greater than or equal to the first change rate, the operating state label is set to a mutation label; When the flow rate change rate is less than the first change rate and greater than or equal to the second change rate, the operating state label is set to a gradual change label; When the flow rate change rate is less than the second change rate, setting the operating state label to a stable label; wherein the first change rate is greater than the second change rate; The operating state labels include sudden change labels, gradual change labels and stable labels.

6. The adaptive integrated pump station dynamic optimization control method according to claim 1, characterized in that: The method of constructing a multi-objective optimization function based on the equipment state coefficient and the working condition state label and generating an alarm signal includes: Construct a multi-objective optimization function; the multi-objective optimization function satisfies the following formula: Among them, j represents the number corresponding to the optimization target, J represents the total number of optimization targets, β j It is expressed as the optimization weight coefficient corresponding to the jth optimization target; the optimization weight coefficient is generated by the equipment status coefficient and the working condition status label; YMF j It is expressed as the target optimization function corresponding to the j-th optimization goal; When the device status coefficient is less than the status threshold, a device damage alarm signal is generated for the device ID corresponding to the device status coefficient; When the operating status tag is a sudden change tag, a water surge alarm signal for the pump station is generated.

7. The adaptive integrated pump station dynamic optimization control method according to claim 6, characterized in that: The optimization weight coefficient is generated by the equipment status coefficient and the working condition status label, including: Extracting the operating status label, equipment status coefficient, electricity price period label, and environmental policy factor; the electricity price period label includes peak electricity price period, valley electricity price period, and flat electricity price period; the environmental policy factor refers to the quantification degree of the current environmental status; Integrate operating status labels, equipment status coefficients, electricity price period labels, and environmental policy factors into optimization analysis data; The optimization analysis data is input into the optimization weight generation model to obtain the optimization weights corresponding to the target optimization functions; the optimization weight generation model is constructed by an artificial intelligence model, including: Obtaining some historical optimization analysis data and their corresponding historical optimization weights; Divide a number of historical optimization analysis data and their corresponding historical optimization weights into training data, verification data, and test data; and perform data preprocessing on the training data, verification data, and test data to obtain a training set, a verification set, and a test set; Select an artificial intelligence model as the base model; Train the basic model using the training set, and adjust the learning rate and other hyperparameters on the validation set to obtain the pre-trained model; By verifying the pre-trained model on the test set, we finally obtain an optimization weight generation model whose input is optimization analysis data and output is the optimization weights corresponding to several target optimization functions.

8. The adaptive integrated pump station dynamic optimization control method according to claim 1, characterized in that: The target optimization function is constructed in the following manner, including: Extract the water inlet flow JL, head YC and operating efficiency YX corresponding to the equipment ID in the equipment parameters; Construct the energy consumption target optimization function NHF, which satisfies the following formula: Where m represents the number corresponding to the device ID, ΔYS represents the device operating time, SM and SG represent the water density and gravity acceleration, respectively; Obtaining electricity prices DJ in different time periods, where the electricity prices are obtained through an electricity price schedule; Extract the device power SP and device operating time ΔYS corresponding to each device ID in different time periods, as well as the number of starts and stops QTC and cumulative operating time LYS corresponding to each device ID; Construct a cost target optimization function CBF, which satisfies the following formula: Where m represents the number corresponding to the device ID, t represents the time period number, c1 and c2 represent the maintenance cost coefficients, c1 represents the cost of a single start and stop, and c2 represents the maintenance cost per unit operating time. Construct the environmental target optimization function HJF, which satisfies the following formula: HJF = tp × NHF; where tp represents the carbon emission factor; The target optimization function includes an energy consumption target optimization function, a cost target optimization function and an environmental target optimization function.

9. The adaptive integrated pump station dynamic optimization control method according to claim 1, characterized in that: The obtaining of control instructions based on the multi-objective optimization function includes: Solving the multi-objective optimization function through a hierarchical optimization strategy to obtain a decision variable; the decision variable refers to the best result that satisfies the multi-objective optimization function under the current state; Parse the decision variables into control set values ​​for each device according to predefined meanings; Map control setting values ​​to control instructions for each device ID.

10. An adaptive integrated pump station dynamic optimization control system, applied to an adaptive integrated pump station dynamic optimization control method according to any one of claims 1 to 9, characterized in that: include: Data acquisition module, data analysis module and control module; The data acquisition module is connected to the data analysis module; the data analysis module is connected to the control module; The data acquisition module is used to acquire equipment parameters, environmental data and pump station data through data acquisition equipment; The data analysis module generates a predicted water inflow based on pump station data and environmental data; generates a device status coefficient based on device parameters; generates a working condition status label based on the predicted water inflow; constructs a multi-objective optimization function based on the device status coefficient and the working condition status label and generates an alarm signal; Obtaining control instructions based on a multi-objective optimization function; The control module dynamically controls related equipment based on control instructions.

Images