A pump station operation method, device and equipment based on time series knowledge graph

Abstract

The present invention relates to the technical field of pump station operation management, and in particular to a pump station operation method, device, equipment and storage medium based on a time series knowledge graph, wherein the method comprises: obtaining real-time monitoring data, and dynamically fitting the pump station unit performance curve based on the real-time monitoring data; obtaining the unit operation feasible domain based on the pump station unit performance curve; constructing a pump station historical operation database, and training a pre-trained model based on the unit operation feasible domain and the pump station historical operation data using cross entropy to obtain a trained model; based on the HyMNet method, using the trained model to predict the current operation plan to obtain a pump station operation plan. The introduced time series dynamic knowledge graph can capture the changes in knowledge over time, and combined with the HyMNet model, effectively use the historical information of the pump station unit operation for prediction, which greatly improves the operational effectiveness of the pump station and ensures the executability and reliability of the optimization strategy.

Description

Technical Field

[0001] The present application relates to the technical field of pump station operation management, and in particular to a pump station operation method, device, equipment and storage medium based on a time series knowledge graph. Background Art

[0002] In terms of pump station operation and management, traditional methods widely rely on empirical rules and static scheduling strategies. These decision-making and planning methods based on historical data and previous experience, although they achieve operation optimization within a certain range, cannot adapt to the complex and changeable actual needs of the site. On the other hand, current advanced attempts mainly focus on using hydraulic models and machine learning methods to find the optimal operation strategy under constraints. These methods usually require precise model parameters and real-time data input, and are committed to finding the optimal strategy for pump station operation at the theoretical level;

[0003] The current technical shortcomings are mainly concentrated in several aspects. First, the traditional data-driven model method is too idealistic in the theoretical construction of the existing scheduling model. For example, the precision of its time period division does not fully take into account the difficulty of actual operation, resulting in a large deviation between the optimization results and the actual operation. Secondly, the accuracy and optimization efficiency of traditional experience and static methods may be seriously affected when faced with the ever-changing performance of pump station units, especially those that change with time, frequency, wear and external environmental factors. When faced with challenges such as changing demands, water level fluctuations and performance changes in actual applications, the adaptability and accuracy of these methods need to be improved.

[0004] In summary, how to design an efficient and energy-saving pump station operation method is an issue that urgently needs to be addressed. Summary of the invention

[0005] The present application aims to solve one of the technical problems in the related art at least to some extent.

[0006] To this end, the first purpose of this application is to propose a pump station operation method based on a time series knowledge graph to solve the problem that the theoretical construction of the scheduling model of the existing technical means is too idealistic, resulting in the optimization results may have a large deviation from the actual operation conditions.

[0007] The second object of the present application is to provide a device.

[0008] The third objective of the present application is to provide an electronic device.

[0009] A fourth objective of the present application is to provide a computer-readable storage medium.

[0010] To achieve the above-mentioned purpose, the first embodiment of the present application proposes a pump station operation method based on a time series knowledge graph, including:

[0011] Acquire real-time monitoring data, and dynamically fit the performance curve of the pump station unit based on the real-time monitoring data;

[0012] Based on the performance curve of the pump station unit, obtaining the feasible operation domain of the unit;

[0013] Constructing a historical operation database of the pump station, and training the pre-trained model using cross entropy based on the feasible operation domain of the unit and the historical operation data of the pump station to obtain a trained model;

[0014] Based on the HyMNet method, the trained model is used to predict the current operation plan to obtain the pump station operation plan.

[0015] Preferably, the acquiring of real-time monitoring data and dynamically fitting the pump station unit performance curve based on the real-time monitoring data comprises:

[0016] Acquire real-time monitoring data, and perform preprocessing operations on the real-time monitoring data to remove outliers to obtain a more accurate fitting effect;

[0017] A cubic polynomial fitting function is used to fit the flow head curve;

[0018] The corresponding function of the flow head curve is obtained based on the polyval function, the three-dimensional graph is fitted using three-dimensional discrete points, and the isoefficiency line is drawn using contour lines.

[0019] Preferably, the calculation formula of the cubic polynomial fitting function is:

[0020]

[0021] in, are the coefficients of the fitted polynomial, is the experimental data involved in curve fitting, is the degree of the fitted polynomial.

[0022] Preferably, obtaining the feasible operation domain of the unit based on the pump station unit performance curve includes:

[0023] Based on the performance curve of the pump station unit, the feasible operation domain of the unit is divided into a stable zone, an economic zone and an unstable zone. During the operation of the pump station, it is necessary to constrain the water supply, blade placement angle, number of units in operation, water level and avoidance conditions.

[0024] Preferably, the construction of the pump station historical operation database is based on the feasible domain of the unit operation and the historical operation data of the pump station, and the pre-trained model is trained using cross entropy to obtain the trained model, which includes:

[0025] Build a data-cleaned historical operation database of the pumping station;

[0026] Using a multi-layer perceptron to access the query vector information, changing the index value of the entity that is not of interest in the N-dimensional vector to a negative number, fusing the query vector information with the historical information, and using a softmax function to estimate the probability of the target entity in the historical vocabulary;

[0027] The pre-trained model is trained using cross entropy to predict a given missing quadruple and obtain a trained model.

[0028] Preferably, the HyMNet method is based on using the trained model to predict the current operation plan, and the pump station operation plan is obtained including:

[0029] The HyMNet method includes two reasoning modes, including:

[0030] History module, which uses historical facts to predict future facts;

[0031] Math module, which uses large-scale system test optimization method to solve the optimal algorithm;

[0032] The solution process can be viewed as predicting a given missing quadruple The historical model will give the predicted object entity with the highest probability in the candidate space, and the mathematical model will give the corresponding theoretical recommendation solution. The HyMNet method will combine the historical model and the mathematical model in a weighted combination to obtain the final prediction value.

[0033] Preferably, the Math module, solving the optimal algorithm by a large system test optimization method, includes:

[0034] Selection of experimental factors, experimental levels and orthogonal arrays to ensure that the experiment covers all possible scenarios and variables;

[0035] Solving the problem of optimal operation of multiple units with variable number in a station using integer programming method;

[0036] Orthogonal analysis was used to determine the theoretically optimal solution.

[0037] To achieve the above-mentioned purpose, the second embodiment of the present application proposes a pump station operation device based on a time series knowledge graph, including:

[0038] A performance curve building module obtains real-time monitoring data and dynamically fits the performance curve of the pump station unit based on the real-time monitoring data;

[0039] A feasible region acquisition module, based on the pump station unit performance curve, acquires the unit operation feasible region;

[0040] A training module, which constructs a historical operation database of the pump station, and uses cross entropy to train the pre-trained model based on the feasible domain of the unit operation and the historical operation data of the pump station to obtain a trained model;

[0041] The prediction module, based on the HyMNet method, uses the trained model to predict the current operation plan to obtain the pump station operation plan.

[0042] To achieve the above-mentioned purpose, the third aspect of the present application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0043] The memory stores computer-executable instructions;

[0044] The processor executes the computer-executable instructions stored in the memory to implement any of the above methods.

[0045] To achieve the above-mentioned purpose, the fourth aspect embodiment of the present application proposes a computer-readable storage medium, including computer-executable instructions stored in the computer-readable storage medium, and the computer-executable instructions are used to implement any of the methods described above when executed by a processor.

[0046] The present application provides a pump station operation method based on a time series knowledge graph. The time series dynamic knowledge graph introduced by the method can capture the changes of knowledge over time. Combined with the HyMNet model, it can effectively use the historical information of the operation of the pump station unit for prediction, and ensure that the optimization plan always captures the latest knowledge changes. Combined with the time series dynamic knowledge graph and the mathematical module, it can take into account the real-time monitoring data and historical scheduling information, ensure that the optimization plan is highly matched with the actual operation scenario, have practical value, greatly improve the operational effectiveness of the pump station, and ensure the executability and reliability of the optimization strategy.

[0047] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0049] Figure 1 A flowchart of a first specific embodiment of a pump station operation method based on a time series knowledge graph provided by the present invention;

[0050] Figure 2 This is an overall flow chart of a pump station optimization operation method based on a time-series dynamic knowledge graph;

[0051] Figure 3 It is a schematic diagram of the performance curve fitting area division;

[0052] Figure 4 This is a schematic diagram of the prediction results of the number of units in operation;

[0053] Figure 5 This is a schematic diagram of the predicted results of the blade placement angle;

[0054] Figure 6 This is the algorithm flow chart of the history module;

[0055] Figure 7 A structural block diagram of a pump station operation device based on a time series knowledge graph provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The core of the present invention is to provide a pump station operation method, device, equipment and storage medium based on a time series knowledge graph. Combined with a time series dynamic knowledge graph and a mathematical module, it can take into account real-time monitoring data and historical scheduling information to ensure that the optimization plan is highly matched with the actual operation scenario, greatly improving the operational effectiveness of the pump station and ensuring the executability and reliability of the optimization strategy.

[0057] In order to enable those skilled in the art to better understand the scheme of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0058] Please refer to Figure 1 , Figure 1 This is a flow chart of a first specific embodiment of a pump station operation method based on a time series knowledge graph provided by the present invention; the specific operation steps are as follows:

[0059] Step S101: acquiring real-time monitoring data, and dynamically fitting a pump station unit performance curve based on the real-time monitoring data;

[0060] Acquire real-time monitoring data, and perform preprocessing operations on the real-time monitoring data to remove outliers to obtain a more accurate fitting effect;

[0061] A cubic polynomial fitting function is used to fit the flow head curve. The calculation formula of the cubic polynomial fitting function is:

[0062]

[0063] in, are the coefficients of the fitted polynomial, is the experimental data involved in curve fitting, is the degree of the fitted polynomial.

[0064] The corresponding function of the flow head curve is obtained based on the polyval function, the three-dimensional graph is fitted using three-dimensional discrete points, and the isoefficiency line is drawn using contour lines.

[0065] Step S102: based on the pump station unit performance curve, obtaining the unit operation feasible region;

[0066] Based on the performance curve of the pump station unit, the feasible operation domain of the unit is divided into a stable zone, an economic zone and an unstable zone. During the operation of the pump station, it is necessary to constrain the water supply, blade placement angle, number of units in operation, water level and avoidance conditions.

[0067] Step S103: constructing a historical operation database of the pump station, and training the pre-trained model using cross entropy based on the feasible domain of the unit operation and the historical operation data of the pump station to obtain a trained model;

[0068] Build a data-cleaned historical operation database of the pumping station;

[0069] Using a multi-layer perceptron to access the query vector information, changing the index value of the entity that is not of interest in the N-dimensional vector to a negative number, fusing the query vector information with the historical information, and using a softmax function to estimate the probability of the target entity in the historical vocabulary;

[0070] The pre-trained model is trained using cross entropy to predict a given missing quadruple and obtain a trained model.

[0071] Step S104: Based on the HyMNet method, the trained model is used to predict the current operation plan to obtain a pump station operation plan.

[0072] The HyMNet method includes two reasoning modes, including:

[0073] History module, which uses historical facts to predict future facts;

[0074] Math module, which uses large-scale system test optimization methods to solve the optimal algorithm, including:

[0075] Selection of experimental factors, experimental levels and orthogonal arrays to ensure that the experiment covers all possible scenarios and variables;

[0076] Solving the problem of optimal operation of multiple units with variable number in a station using integer programming method;

[0077] Orthogonal analysis was used to determine the theoretically optimal solution.

[0078] The solution process can be viewed as predicting a given missing quadruple The historical model will give the predicted object entity with the highest probability in the candidate space, and the mathematical model will give the corresponding theoretical recommendation solution. The HyMNet method will combine the historical model and the mathematical model in a weighted combination to obtain the final prediction value.

[0079] This embodiment provides a pump station operation method based on a time series knowledge graph, which uses real-time monitoring data to dynamically fit the unit performance and generate a more accurate and time-effective performance curve. It ensures the efficient and stable operation of the pump station, helps to extend the service life of the unit and reduce maintenance costs. The introduced time series dynamic knowledge graph can capture the changes in knowledge over time. Combined with the HyMNet model, it effectively uses the historical information of the pump station unit operation for prediction, ensuring that the optimization plan always captures the latest knowledge changes.

[0080] Based on the above embodiment, this embodiment describes the pump station operation method based on the time series knowledge graph. Figure 2 As shown, the details are as follows:

[0081] like Figure 3 As shown, the performance curve of the pump station unit is dynamically fitted using real-time monitoring data to provide basic data for subsequent steps;

[0082] First, the data is preprocessed to remove outliers to obtain a more accurate fitting effect;

[0083] A cubic polynomial fitting function is used to fit the flow head curve. Excluding high-order polynomials will cause the Rung phenomenon. The function returns the coefficients of the fitted polynomial (arranged in descending order). is the experimental data involved in curve fitting, To fit the degree of the polynomial, the formula is as follows:

[0084]

[0085] Depend on The function is used to find the function corresponding to the curve fitted by S102;

[0086] The 3D graph is fitted with 3D discrete points, and the contour lines are further used to draw the isoefficiency lines. Constructing a three-dimensional coordinate system ;

[0087] First, use MeshGrid combined with min and max functions to generate grid coordinates, and then use GridData function to interpolate the Z values ​​of these points. The interpolation method is cubic to automatically eliminate points outside the measurement range;

[0088] Finally, use MATLAB's surf function to draw a three-dimensional graph. After creating a three-dimensional surface plot to represent the three-dimensional matrix data, use The function generates a contour plot. Indicates the level of the contour line. The formula is as follows:

[0089]

[0090]

[0091] Based on the dynamic performance curve, combined with mathematical models and decision analysis, the feasible region of unit operation is determined;

[0092] Divided into stable zone, economic zone and unstable zone;

[0093] Defining Deviation It is the ratio of the difference between the working point flow rate and the best efficiency point flow rate at the current blade angle of the pump to the best efficiency point flow rate. It is Traffic corresponding to the time period , For the The flow rate corresponding to the best efficiency point in the time period The formula is as follows:

[0094]

[0095] The criterion for determining the stable region is that the reliability is range, otherwise it is an unstable area;

[0096] The goal is economic optimization, where The minimum power consumption cost of the pump station per day, 10,000 yuan; The reliability of the pump; Indicates pump station The number of machines in operation during the time period; is the water density, ; is the acceleration due to gravity, ; For the The average lift during the period, are the pump unit efficiency and the motor efficiency respectively; With It is related to the flow rate and head during the period; for The electricity price of the time period is defined as a constant in this paper because the price of electricity in a single day does not change much. The formula is as follows:

[0097]

[0098] On the basis of ensuring stability, determine the economic operation zone. Define the deviation It is the ratio of the difference between the working point flow rate and the most economical point flow rate at the current blade angle of the pump to the most economical point flow rate. For the The flow rate corresponding to the most economical point in the time period is as follows:

[0099]

[0100] The judgment condition of economic zone is that the deviation is The opposite is the non-economic area.

[0101] The constraints required during the operation of the pump station include water supply, blade placement angle, number of units in operation, water level, and avoidance conditions;

[0102] Constraints on the number of machines started, is the maximum number of pumps that can be operated in the pumping station, such as Figure 4 As shown, if a unit failure occurs, the maximum number of units is dynamically planned, and the formula is as follows:

[0103]

[0104] Pump station water supply constraints, pump station water supply includes total flow and head constraints. The total amount of water pumped by the pump station during daily operation. For the The length of the time period is as follows:

[0105]

[0106] Water level constraints, taking into account ecological water level / waterway water level / flood control water level. , They are the minimum and maximum blade placement angles allowed for adjustment of the units in the pump station, respectively. The formula is as follows:

[0107]

[0108] Unit power constraints (head constraints / flow constraints), For the The actual operating power of the unit when the blade placement angle is, The rated operating power of the motor is as follows:

[0109]

[0110] like Figure 5 As shown, the blade placement angle constraint is , They are the minimum and maximum blade placement angles allowed for adjustment of the units in the pump station, respectively. The formula is as follows:

[0111]

[0112] Avoid operating constraints and determine special vibration areas through field testing / long-term observation, such as the saddle area of ​​the head curve of the axial flow pump and the low flow operating area. , They are the minimum allowable pumping volume of a single unit (water pump) in a pumping station. The formula is as follows:

[0113]

[0114] Build a historical operation database of the pump station, integrate historical information and query information, and use cross entropy to train the model to provide historical reference for subsequent predictions;

[0115] First, a detailed and data-cleaned historical database of pump station operations is established to provide a basis for subsequent steps. is an N-dimensional vector (N is the total number of entities). An element of 1 in the vector indicates that there is a fact with the corresponding entity as the objective in the training set, and the time is , otherwise it is 0;

[0116] Summarize historical information. The formula is as follows:

[0117]

[0118] Then an MLP is used to access the query vector information for subsequent processing. is the unit step of time, and let ; Index vector is an N-dimensional vector, where N is the entire entity library The base number, and Model training parameters; are all embedding vectors. The formula is as follows:

[0119]

[0120]

[0121] In order to reduce the probability of useless information, The index values ​​of entities that are not of interest in become negative, and we get Finally, the query information is fused with the history information, and the softmax function is used to estimate the probability of the target entity in the historical vocabulary. is a vector of length N, representing the predicted probability on the historical vocabulary. Each element represents the probability that the predicted object is the corresponding (index number) entity. The formula is as follows:

[0122]

[0123] The solution process is considered as predicting a given missing quadruple , cross entropy will be used in the model for model training. Is a snapshot The A real object entity, is when the real object entity is Snapshot Middle The combined probability value of object entities. The formula is as follows:

[0124]

[0125] The HyMNet method is proposed, which combines the history module and the mathematical module to make real-time predictions on the current operation plan, thereby selecting the optimal entity facts;

[0126] A TKG representation learning method based on a novel time-aware replication mechanism, HyMNet (Temporal History-Math Network), can not only predict future facts from the entire entity vocabulary, but also repeatedly identify facts and select them based on the historical vocabulary that formed the entities that appeared in the past. HyMNet contains two reasoning modes: History module and Math module;

[0127] The History module uses historical facts to predict future facts. A History module is built based on the time-series dynamic knowledge graph, such as Figure 6 As shown, historical data is used for prediction;

[0128] Math module is to solve the optimal algorithm through large system test optimization method;

[0129] The solution process can be viewed as predicting a given missing quadruple , the historical model will give the predicted object entity with the highest probability in the candidate space, and the mathematical model will give the corresponding theoretical recommendation solution. HyMNet will combine the historical model and the mathematical model to get the final prediction value. Finally, the administrator will choose the economic or stable solution based on his or her personality. Introduce a similarity parameter , depends on the deviation of the target operating range, that is, the proportion of the absolute value of the difference between the actual data and the target data to the target data. is the entity library vector containing all entity probabilities, and the formula is as follows:

[0130]

[0131] Over time, the HyMNet model is used to provide optimal recommendations for pumping stations regarding start-up time, number of running units, and blade angles, ensuring efficient and stable operation of the pumping stations.

[0132] Through the above method, the performance curve is dynamically updated and the operation area is divided according to the real-time monitoring data;

[0133] Organize historical data, ensure the consistency of its format and the integrity of the data, and then divide the data into training sets and test sets;

[0134] As new data is constantly generated, relying solely on early historical data may not be able to capture the latest trends and patterns. Therefore, dynamically fill in new data and expand the History module to ensure that the model is always trained based on the latest and most comprehensive data, improving the accuracy and reliability of its predictions;

[0135] Based on HyMNet, the optimal startup time, number of units and corresponding blade placement angle are recommended.

[0136] The pump station operation method based on the time series knowledge graph provided by the embodiment of the present invention is highly accurate and real-time. Especially in long-distance water diversion projects, its application can ensure that the pump station can achieve more accurate and intelligent optimization operation in complex environments, and dynamically fit the unit performance using real-time monitoring data to generate more accurate and timely performance curves. To ensure the efficient and stable operation of the pump station, and help extend the service life of the unit and reduce maintenance costs, this method provides a comprehensive decision-making framework, using mathematical models to divide the operation area in multiple dimensions, such as "economic operation" and "stable operation", to ensure that the optimization strategy is both comprehensive and practical. The introduced time series dynamic knowledge graph can capture the changes in knowledge over time, and combined with the HyMNet model, effectively use the historical information of the pump station unit operation for prediction, to ensure that the optimization plan always captures the latest knowledge changes.

[0137] Please refer to Figure 7 , Figure 7 A structural block diagram of a pump station operation device based on a time series knowledge graph provided in an embodiment of the present invention; the specific device may include:

[0138] The performance curve building module 100 acquires real-time monitoring data and dynamically fits the performance curve of the pump station unit based on the real-time monitoring data;

[0139] A feasible region acquisition module 200, based on the pump station unit performance curve, acquires the unit operation feasible region;

[0140] The training module 300 constructs a historical operation database of the pump station, and trains the pre-trained model using cross entropy based on the feasible domain of the unit operation and the historical operation data of the pump station to obtain a trained model;

[0141] The prediction module 400 predicts the current operation plan based on the HyMNet method using the trained model to obtain the pump station operation plan.

[0142] A pump station operation device based on a time series knowledge graph in this embodiment is used to implement the aforementioned pump station operation method based on a time series knowledge graph. Therefore, the specific implementation method of a pump station operation device based on a time series knowledge graph can be seen in the embodiment part of a pump station operation method based on a time series knowledge graph in the previous text. For example, the performance curve construction module 100, the feasible domain acquisition module 200, the training module 300, and the prediction module 400 are respectively used to implement steps S101, S102, S103, and S104 in the aforementioned pump station operation method based on a time series knowledge graph. Therefore, its specific implementation method can refer to the description of the corresponding embodiments of each part, which will not be repeated here.

[0143] In order to implement the above embodiments, the present application also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method provided by the above embodiments.

[0144] In order to implement the above embodiments, the present application also proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0145] In order to implement the above embodiments, the present application also proposes a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0146] The collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in this application are in compliance with relevant laws and regulations and do not violate public order and good morals.

[0147] It should be noted that personal information from users should be collected for legitimate and reasonable purposes and should not be shared or sold outside of these legitimate uses. In addition, such collection / sharing should be carried out after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign the agreement / authorization including authorization of relevant user information before the user uses the function. In addition, any necessary steps should be taken to protect and safeguard access to such personal information data and ensure that others who have access to personal information data comply with its privacy policy and procedures.

[0148] The present application is expected to provide an implementation scheme for users to selectively block the use or access of personal information data. That is, the present disclosure is expected to provide hardware and / or software to prevent or block access to such personal information data. Once the personal information data is no longer needed, the risk can be minimized by limiting data collection and deleting the data. In addition, when applicable, such personal information is de-identified to protect the privacy of the user.

[0149] In the description of the aforementioned embodiments, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they contradict each other.

[0150] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0151] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0152] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0153] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0154] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0155] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0156] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A pump station operation method based on a time series knowledge graph, characterized in that: include: Acquire real-time monitoring data, and dynamically fit the performance curve of the pump station unit based on the real-time monitoring data, wherein the real-time monitoring data is preprocessed to remove abnormal values ​​to obtain a more accurate fitting effect, a cubic polynomial fitting function is used to fit the flow head curve, and a corresponding function of the flow head curve is obtained based on the polyval function, a three-dimensional discrete point is used to fit a three-dimensional graph, and an isoefficiency line is drawn using contour lines; Based on the performance curve of the pump station unit, the feasible domain of the unit operation is obtained, wherein the feasible domain of the unit operation is divided into a stable zone, an economic zone and an unstable zone based on the performance curve of the pump station unit. During the operation of the pump station, it is necessary to constrain the water supply, blade placement angle, number of units in operation, water level and avoidance conditions; Constructing a historical operation database of a pump station, and based on the feasible domain of the unit operation and the historical operation data of the pump station, using cross entropy to train the pre-trained model to obtain a trained model, wherein a historical operation database of a pump station after data cleaning is constructed, and a multi-layer perceptron is used to access the query vector information, and the index value of an entity that is not of interest in the N-dimensional vector is changed to a negative number, and the query vector information is fused with the historical information, and a softmax function is used to estimate the probability of the target entity in the historical vocabulary, and the pre-trained model is trained using cross entropy to predict a given missing quadruple to obtain a trained model; Based on the HyMNet method, the trained model is used to predict the current operation plan to obtain the pump station operation plan, wherein the HyMNet method includes two reasoning modes including: History module, which uses historical facts to predict future facts; Math module, which uses large-scale system test optimization method to solve the optimal algorithm; The solution process is regarded as predicting a given missing quadruple (s, p, ?, t). The historical model will give the predicted object entity with the highest probability in the candidate space, and the mathematical model will give the corresponding theoretical recommendation solution. The HyMNet method will combine the historical model and the mathematical model in a weighted combination to obtain the final prediction value.

2. The pump station operation method based on time series knowledge graph according to claim 1 is characterized in that: The cubic polynomial fitting function calculation formula is: p=ployfit(X,Y,n) Among them, p is the coefficient of the fitting polynomial, X and Y are the experimental data involved in curve fitting, and n is the degree of the fitting polynomial.

3. The pump station operation method based on time series knowledge graph according to claim 1 is characterized in that: The Mathmodule uses a large system test optimization method to solve the optimal algorithm solution including: Selection of experimental factors, experimental levels and orthogonal arrays to ensure that the experiment covers all possible scenarios and variables; Solving the problem of optimal operation of multiple units with variable number in a station using integer programming method; Orthogonal analysis was used to determine the theoretically optimal solution.

4. A pump station operation device based on a time series knowledge graph, characterized in that: include: A performance curve construction module obtains real-time monitoring data, and dynamically fits the performance curve of the pump station unit based on the real-time monitoring data, wherein the real-time monitoring data is preprocessed to remove abnormal values ​​to obtain a more accurate fitting effect, a cubic polynomial fitting function is used to fit the flow head curve, and the corresponding function of the flow head curve is obtained based on the polyval function, three-dimensional discrete points are used to fit the three-dimensional graph, and contour lines are used to draw isoefficiency lines; A feasible domain acquisition module is used to acquire the feasible domain of the unit operation based on the performance curve of the pump station unit, wherein the feasible domain of the unit operation is divided into a stable area, an economic area and an unstable area based on the performance curve of the pump station unit, and the water supply, blade placement angle, number of units in operation, water level and avoidance conditions need to be constrained during the operation of the pump station; A training module, constructing a historical operation database of a pump station, and based on the feasible domain of the unit operation and the historical operation data of the pump station, using cross entropy to train the pre-trained model to obtain a trained model, wherein a historical operation database of a pump station after data cleaning is constructed, query vector information is accessed using a multi-layer perceptron, entity index values ​​of no interest in the N-dimensional vector are changed to negative numbers, the query vector information is fused with historical information, a softmax function is used to estimate the probability of a target entity in a historical vocabulary, and cross entropy is used to train the pre-trained model to predict a given missing quadruple to obtain a trained model; The prediction module, based on the HyMNet method, uses the trained model to predict the current operation plan to obtain the pump station operation plan, wherein the HyMNet method includes two reasoning modes including: History module, which uses historical facts to predict future facts; Math module, which uses large-scale system test optimization method to solve the optimal algorithm; The solution process is regarded as predicting a given missing quadruple (s, p, ?, t). The historical model will give the predicted object entity with the highest probability in the candidate space, and the mathematical model will give the corresponding theoretical recommendation solution. The HyMNet method will combine the historical model and the mathematical model in a weighted combination to obtain the final prediction value.

5. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 3 when executed by a processor.

Images