Remote control system for water pump for production
By designing a remote control system, using wireless communication technology to collect and transmit pump station pressure data in real time and generate control instructions, the problems of inconvenience of communication between the pump station and the power plant and waste of human resources are solved, and the automation control and management efficiency of the pump station are improved .
Patent Information
- Application Number
- CN202410991266.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-23
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The distance between the pump station and the power plant is relatively long, which leads to inconvenient communication, high labor control costs, long maintenance and management response time, making it difficult to achieve automated control.
A remote control system for production water pumps is designed, including cameras, hard disk burners, wireless routers, central video hosts, central control hosts, radio transceivers and receiving equipment and pump station control cabinets, and the pump water inlets are collected and transmitted in real time through wireless communication. and the outlet pressure gauge image, extract pressure values and generate control instructions to realize remote control and monitoring of the pump.
Unattended pump station management is realized, the management operation costs are reduced, maintenance response time is shortened, the management level of pump stations is improved, and the problems of communication inconvenience and human resources waste are solved through automated control.
Smart Images

Figure CN118934584A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of pump station control, and in particular relates to a remote control system for a production water pump. Background Art
[0002] The power plant water supply pump station is an important engineering facility, usually composed of production water pumps, motors, pipelines, valves and control systems, so as to provide the power plant with the required production water and ensure the normal operation of the power plant.
[0003] At present, the long distance between the pump station and the power plant leads to the following defects: (1) It is inconvenient for the pump station and the power plant to communicate with each other, and problems cannot be discovered and handled in time; (2) If the traditional method of dedicated personnel taking turns on duty day and night is adopted, a large amount of human resources will be wasted and it will not be conducive to information management; (3) It is very inconvenient to construct and wire the transmission line, which makes it difficult to solve the communication inconvenience problem even if wired transmission is used.
[0004] Therefore, in order to effectively reduce the management and operation costs of pump stations, shorten the maintenance and management response time of pump stations, reduce the equipment failure rate of pump stations, improve the company's management level of pump stations, and realize the remote security function of pump stations, it is necessary to provide a new solution for remote control and monitoring of production water pumps. Summary of the invention
[0005] The purpose of the present invention is to provide a remote control system for production water pumps to solve the existing problems of inconvenient communication, high human control cost, long maintenance and management response time and difficulty in achieving automated control due to the long distance between the pump station and the power plant.
[0006] In order to achieve the above object, the present invention adopts the following technical solutions:
[0007] In the first aspect, a remote control system for a production water pump is provided, including a camera, a hard disk recorder, a wireless router, a central video host, a central control host, a radio station transceiver and a pump station control cabinet;
[0008] The camera is arranged in the target pump station and connected to the hard disk recorder by wired communication, and is used to collect the image of the water inlet pressure gauge and the image of the water outlet pressure gauge of the production water pump in the target pump station in real time, and transmit the image of the water inlet pressure gauge and the image of the water outlet pressure gauge to the hard disk recorder in real time;
[0009] The hard disk burner is arranged in the target pump station and connected to the wireless router via wired communication, and is used to burn the received water inlet pressure gauge image and the water outlet pressure gauge image into the local hard disk in real time, and forward the water inlet pressure gauge image and the water outlet pressure gauge image to the wireless router in real time;
[0010] The wireless router is wirelessly connected to the central video host and is used to forward the received water inlet pressure gauge image and the water outlet pressure gauge image to the central video host in real time;
[0011] The central video host is connected to the central control host by wired communication, and is used to extract the water inlet pressure value and the water outlet pressure value of the production water pump in real time from the received water inlet pressure gauge image and the water outlet pressure gauge image, and transmit the water inlet pressure value and the water outlet pressure value to the central control host in real time;
[0012] The central control host is connected to the radio station transceiver device via wired communication, and is used to generate a control instruction for starting the production water pump, stopping the production water pump, or adjusting the operation frequency of the production water pump after starting according to the received water inlet pressure value and the water outlet pressure value, and send the control instruction to the radio station transceiver device in real time;
[0013] The radio station transceiver device is wirelessly connected to the pump station control cabinet and is used to forward the received control instructions to the pump station control cabinet in real time;
[0014] The pump station control cabinet is arranged in the target pump station and is connected to the controlled end of the production water pump by wired communication, and is used to execute the received control instructions in real time to start the production water pump, stop the production water pump or adjust the operating frequency of the production water pump after starting.
[0015] Based on the above invention content, a remote control solution for production water pumps for achieving unattended operation is provided, which includes a camera, a hard disk recorder, a wireless router, a central video host, a central control host, a radio station transceiver and a pump station control cabinet. Through their communication connection relationship and collaboration, the images of the water inlet pressure gauge and the water outlet pressure gauge of the production water pump can be collected in real time, and then the water inlet pressure value and the water outlet pressure value of the water pump can be extracted in real time from these images. Then, according to these pressure values, control instructions for starting and stopping the water pump or adjusting the operating frequency of the water pump after starting are generated in real time, and finally the control instructions are wirelessly sent to the pump station control cabinet for execution. In this way, not only can the communication inconvenience between the power plant and the pump station be solved based on wireless communication technology, but there is no need to send people to take turns on duty day and night, and it is conducive to information management, thereby realizing the purpose of automatic control of the pump station, effectively reducing the management and operation costs of the pump station, shortening the response time of the pump station maintenance management, and improving the management level of the pump station by the enterprise, which is convenient for practical application and promotion.
[0016] In a possible design, the camera includes a water inlet side camera and a water outlet side camera respectively connected to the hard disk recorder by wired communication;
[0017] The water inlet side camera is arranged around the water inlet of the production water pump and allows the camera field of view to cover the water inlet pressure gauge of the production water pump, and is used to collect the image of the water inlet pressure gauge of the production water pump in real time, and transmit the image of the water inlet pressure gauge to the hard disk recorder in real time;
[0018] The outlet side camera is arranged around the outlet of the production water pump and makes the lens field of view cover the outlet pressure gauge of the production water pump. It is used to collect the outlet pressure gauge image of the production water pump in real time and transmit the outlet pressure gauge image to the hard disk burner in real time.
[0019] In one possible design, the camera also includes a field monitoring camera connected to the hard disk recorder via wired communication;
[0020] The on-site monitoring camera is arranged around the production water pump and covers the production water pump in its field of view, and is used to collect on-site monitoring images of the production water pump in real time and transmit the on-site monitoring images to the hard disk recorder in real time;
[0021] The hard disk burner is also used to burn the received on-site monitoring images into a local hard disk in real time, and forward the on-site monitoring images to the wireless router in real time;
[0022] The wireless router is also used to forward the received on-site monitoring images to the central video host in real time;
[0023] The central video host is also used to display the received on-site monitoring images in real time.
[0024] In a possible design, the central control host is also used to send the control instruction to the central video host in real time;
[0025] The central video host is also used to determine whether the working status of the target pumping station is consistent with the control instruction based on the received on-site monitoring image after receiving the control instruction. If not, trigger the execution of a fault alarm action for the target pumping station.
[0026] In a possible design, if the control instruction is used to start the production water pump or stop the production water pump, judging whether the working state of the target pump station is consistent with the control instruction according to the received on-site monitoring image includes:
[0027] Using a target detection algorithm to perform production water pump image recognition processing on the received on-site monitoring image to obtain a water pump image recognition result;
[0028] According to the production water pump mark frame in the water pump image recognition result, an image of the production water pump is intercepted from the on-site monitoring image;
[0029] Importing the production water pump image into a water pump state recognition model pre-trained based on a convolutional neural network, and outputting a water pump state recognition result;
[0030] It is determined whether the water pump state identification result matches the control instruction. If so, it is determined that the working state of the target pump station matches the control instruction. Otherwise, it is determined that the working state of the target pump station does not match the control instruction.
[0031] In a possible design, if the control instruction is used to adjust the operating frequency of the production water pump after startup, judging whether the working state of the target pump station is consistent with the control instruction according to the received on-site monitoring image includes:
[0032] Using a target detection algorithm to perform production water pump image recognition processing on the received on-site monitoring image to obtain a water pump image recognition result;
[0033] According to the production water pump mark frame in the water pump image recognition result, an image of the production water pump is intercepted from the on-site monitoring image;
[0034] Extracting from the production water pump image a current operating frequency displayed in a working parameter display screen of the production water pump;
[0035] It is determined whether the current operating frequency matches the target operating frequency in the control instruction. If so, it is determined that the working state of the target pump station matches the control instruction. Otherwise, it is determined that the working state of the target pump station does not match the control instruction.
[0036] In a possible design, when the target pump station is a primary pump station, a control instruction for starting the production water pump, stopping the production water pump, or adjusting the operation frequency of the production water pump after starting is generated in real time according to the received water inlet pressure value and the water outlet pressure value, including:
[0037] Acquire water inlet pressure time series data, wherein the water inlet pressure time series data contains a plurality of intra-day unit time periods that are sequentially continuous in time series and a historical water inlet pressure value of each intra-day unit time period in the plurality of intra-day unit time periods;
[0038] Dividing the water inlet pressure time series data into a plurality of water inlet pressure time series data corresponding one to one to a plurality of time windows that are sequentially continuous in time series;
[0039] For the last time window among the multiple time windows, at least one piece of water inlet pressure time series data corresponding to at least one corresponding previous time window is used as a corresponding training data set, and a piece of water inlet pressure time series data corresponding to the corresponding window is used as a corresponding verification data set;
[0040] Based on the Prophet model, a real-time dynamic prediction model for water inlet pressure is established using the following formula:
[0041]
[0042] Wherein, t represents time, A(t) represents the water inlet pressure value at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, l q (t) represents the linear regression term corresponding to the qth linear influence factor among the Q linear influence factors at time t;
[0043] The growth mode of the trend term is set to a linear growth mode, and natural weeks and natural days are added to make the seasonal term change periodically, and the error noise term is also set to obey a mathematical expectation of μ and a variance of σ 2 Normal distribution of
[0044] Acquire the time series data of all water inlet pressure related influencing factors, and then dynamically screen the water inlet pressure linear influencing factors for all the water inlet pressure related influencing factors based on the time series data, so as to determine the water inlet pressure related influencing factors that meet the screening conditions as the linear influencing factors, and obtain a model parameter vector composed of weight adjustment parameters in the linear regression items of all the linear influencing factors;
[0045] Using the training data set and the verification data set corresponding to the last time window, the real-time dynamic prediction model of the water inlet pressure is trained and optimized based on the particle swarm algorithm to obtain a verified model parameter optimization search result, wherein the model parameter optimization search result includes the optimization search result of the model parameter vector, the mathematical expectation and the variance;
[0046] Inputting the search results of the model parameter optimization into the real-time dynamic prediction model for water inlet pressure, and then applying the real-time dynamic prediction model for water inlet pressure to predict the predicted value of water inlet pressure in a unit time period within the current day, a unit time period within the next day, or a unit time period within at least two days in the future;
[0047] According to a plurality of water inlet pressure value ranges that are preset and correspond one to one with a plurality of water pump energy-saving frequency conversion strategies, a certain water inlet pressure value range that includes the water inlet pressure prediction value is determined from the plurality of water inlet pressure value ranges;
[0048] According to a water pump energy-saving and frequency-changing strategy among the multiple water pump energy-saving and frequency-changing strategies and corresponding to the certain water inlet pressure value range, a control instruction for starting the production water pump, stopping the production water pump or adjusting the operating frequency of the production water pump after starting is generated in real time according to the received water inlet pressure value and the water outlet pressure value.
[0049] In a possible design, the linear influencing factors of the water inlet pressure are dynamically screened for all the inlet pressure-related influencing factors based on the time series data, so as to determine the inlet pressure-related influencing factors that meet the screening conditions as the linear influencing factors, including:
[0050] Randomly select historical water inlet pressures of multiple natural days to form a first sample S1, and select data of the multiple natural days from the corresponding time series data as a second sample S2 for a certain water inlet pressure-related influencing factor among all the water inlet pressure-related influencing factors;
[0051] Performing a normal distribution KS check on the first sample S1 to calculate a first check statistic p value p1, and also performing the normal distribution KS check on the second sample S2 to calculate a second check statistic p value p2;
[0052] If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the first preset threshold, the correlation coefficient r between the first sample S1 and the sample S2 is calculated according to the following formula: q :
[0053]
[0054] In the formula, i represents a positive integer, x i represents the i-th sample value in the first sample S1, y i represents the i-th sample value in the second sample S2, represents the sample mean of the first sample S1, represents the sample mean of the second sample S2;
[0055] If the correlation coefficient r q If the inlet pressure is greater than a second preset threshold, it is determined that the certain inlet pressure-related influencing factor meets the screening condition, and the certain inlet pressure-related influencing factor is determined as the linear influencing factor.
[0056] In a possible design, the training data set and the verification data set corresponding to the last time window are used to perform model training and parameter optimization on the water inlet pressure real-time dynamic prediction model based on the particle swarm algorithm to obtain a verified model parameter optimization search result, including the following steps S701 to S712:
[0057] S701. Initialize the algorithm parameters of the particle swarm algorithm including the population size N, the maximum number of iterations T, the self-learning factor C1 and the global learning factor C2, and randomly generate the initial position of each particle in the particle population, and then execute step S702, wherein the initial position of each particle is randomly generated based on the following formula:
[0058]
[0059] Wherein, i' represents a positive integer, d represents a positive integer less than or equal to D, D represents the number of dimensions of the model parameter vector to be optimized, and the model parameter vector to be optimized includes the model parameter vector to be optimized, the Mathematical Expectation and stated variance , represents the initial position of the i'th particle in the particle population on the dth dimension of the model parameter vector to be optimized, u c,d represents the upper limit of the model parameter search space in the dth dimension, l c,d represents the lower limit of the model parameter search space in the d-th dimension, and rand(0,1) represents a pure decimal random generator function;
[0060] S702. For each particle, the corresponding initial position is input into the water inlet pressure real-time dynamic prediction model as the model parameter search result, and then the training data set corresponding to the last time window is used to train the water inlet pressure real-time dynamic prediction model, and the corresponding fitness F is calculated, and then step S703 is executed, wherein the fitness F is calculated according to the following formula:
[0061]
[0062] Where n represents the total number of samples in the training data set, i represents a positive integer, represents the prediction output item calculated after the model input item of the i-th sample data in the training data set is imported into the water inlet pressure real-time dynamic prediction model, y (i) represents the true output item of the i-th sample data;
[0063] S703. Take the initial position of a particle with the minimum fitness as the initial global optimal position G ini , the position of the i'th particle in the iteration process and corresponding to the minimum fitness is the local optimal position P reached by the i'th particle i’ , and the position of the particle population in the iteration process and corresponding to the minimum fitness is recorded as the global optimal position G reached by the particle population, and the moving speed of the i'th particle is recorded as V i’ , and finally initialize the local optimal position P i’ =G=G ini , and initialize the moving speed V i’ is zero, and the particle movement inertia coefficient range is also set to [T min ,T max ], and initialize the period T of the periodic reduction of the inertia coefficient to a randomly selected value within the range of the particle movement inertia coefficient, and also initialize the particle movement inertia coefficient in the iteration process to a value greater than zero, and also initialize the current iteration number t' to 0, and then execute step S704;
[0064] S704. Update the particle movement inertia coefficient of each particle, and then execute step S705, wherein the update formula of the particle movement inertia coefficient is as follows:
[0065]
[0066] In the formula, represents the particle movement inertia coefficient of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the particle movement inertia coefficient of the i'th particle in the dth dimension before the t'+1th iteration, t'%T represents the modulo operation result of the previous iteration number t' on the period T, and if If it is less than 0, a new value is randomly selected within the range of the particle's inertia coefficient as the new period T, and the calculation is recalculated. until Not less than 0;
[0067] S705. Update the moving speed of each particle, and then execute step S706, wherein the updating formula of the moving speed is as follows:
[0068]
[0069] In the formula, represents the moving speed of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the moving speed of the i'th particle in the dth dimension before the t'+1th iteration, γ1 represents the first random number uniformly distributed in (0,1), represents the local optimal position reached by the i'th particle in the dth dimension during the first t' iterations, represents the position of the i'th particle in the dth dimension before the t'+1th iteration, γ2 represents a second random number uniformly distributed in (0,1), represents the global optimal position reached by the particle population in the d-th dimension during the first t' iterations;
[0070] S706. Update the position of each particle, and then execute step S707, wherein the update formula of the position is as follows:
[0071] In the formula, represents the position of the i'th particle in the dth dimension obtained after the t'+1th iteration;
[0072] S707. For each particle, the corresponding position is input into the water inlet pressure real-time dynamic prediction model as the model parameter search result, and then the training data set corresponding to the last time window is used again to perform model training on the water inlet pressure real-time dynamic prediction model, and the corresponding fitness F is calculated again, and then step S708 is executed;
[0073] S708. Determine the local optimal position P i’ Is the corresponding fitness greater than the current fitness of the i'th particle? If so, the local optimal position P i’ Update to the current position of the i'th particle, and then execute step S709, otherwise directly execute step S709;
[0074] S709. Determine whether the fitness corresponding to the global optimal position G is greater than the minimum value among the current fitness of each particle. If so, update the global optimal position G to the current position of any particle with the minimum value, and then execute step S710. Otherwise, directly execute step S710;
[0075] S710. Increment the current number of iterations t' by 1, and determine whether the current number of iterations t' reaches the maximum number of iterations T. If so, execute step S711, otherwise return to execute step S704;
[0076] S711. Input the global optimal position G as the model parameter search result into the water inlet pressure real-time dynamic prediction model, and then use the water inlet pressure real-time dynamic prediction model and the verification data set corresponding to the last time window to predict the water inlet pressure, and calculate the new fitness F new , then execute step S712;
[0077] S712. Judge F new <ρ×F min Is it true? If so, the global optimal position G is used as the verified model parameter optimization search result. Otherwise, after adjusting the search space of the model parameter vector to be optimized, return to step S701, where ρ represents a preset error verification coefficient and is greater than 1, and F min represents the fitness corresponding to the global optimal position G, and the model parameter optimization search result includes the model parameter vector, the Mathematical Expectation and stated square Difference Optimize search results.
[0078] In a possible design, all the inlet pressure-related influencing factors include wind level, wind speed, time to sunrise on the same day, time to sunset on the same day, cloud cover, perceived temperature, visibility, precipitation probability, humidity, dew point temperature, western Pacific subtropical high intensity index value, Indo-Myanmar trough intensity index value, Tibet Plateau-1 index value, tropical Indian Ocean sea surface temperature dipole index value, subtropical South Indian Ocean dipole index value, NINO 3.4 region sea surface temperature anomaly index value and / or sunspot index value, wherein the current day refers to the day corresponding to the unit time period within the day.
[0079] Beneficial effects of the above scheme:
[0080] (1) The present invention creatively provides a remote control solution for production water pumps for achieving the purpose of unattended operation, which includes a camera, a hard disk recorder, a wireless router, a central video host, a central control host, a radio station transceiver and a pump station control cabinet. Through their communication connection relationship and cooperation, the images of the water inlet pressure gauge and the water outlet pressure gauge of the production water pump can be collected in real time, and then the water inlet pressure value and the water outlet pressure value of the water pump can be extracted in real time from these images. Then, according to these pressure values, control instructions for starting and stopping the water pump or adjusting the operating frequency of the water pump after starting are generated in real time. Finally, the control instructions are wirelessly sent to the pump station control cabinet for execution. In this way, not only can the communication inconvenience between the power plant and the pump station be solved based on wireless communication technology, but there is no need to send people to take turns on duty day and night, and it is conducive to information management, thereby realizing the purpose of automatic control of the pump station, effectively reducing the management and operation costs of the pump station, shortening the response time of the pump station maintenance management, and improving the management level of the pump station by the enterprise, which is convenient for practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS
[0081] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0082] Figure 1 This is a schematic diagram of the structure of a remote control system for a production water pump provided in an embodiment of the present application. DETAILED DESCRIPTION
[0083] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.
[0084] It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another object. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.
[0085] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.
[0086] Example
[0087] like Figure 1 As shown, the production water pump remote control system provided in this embodiment includes but is not limited to a camera 1, a hard disk burner 2, a wireless router 3, a central video host 4, a central control host 5, a radio station transceiver 6 and a pump station control cabinet 7.
[0088] The camera 1 is arranged in the target pump station and connected to the hard disk recorder 2 by wired communication, and is used to collect the images of the water inlet pressure gauge and the water outlet pressure gauge of the production water pump in the target pump station in real time, and transmit the images of the water inlet pressure gauge and the water outlet pressure gauge to the hard disk recorder 2 in real time. Figure 1 As shown, the target pump station can be a primary pump station in a two-stage pump station pressurized conveying system, or a secondary pump station in the two-stage pump station pressurized conveying system. Specifically, the camera 1 includes but is not limited to an inlet side camera 11 and an outlet side camera 12 respectively connected to the hard disk recorder 2 by wired communication; the inlet side camera 11 is arranged around the inlet of the production water pump and the lens field of view covers the inlet pressure gauge of the production water pump, and is used to collect the image of the inlet pressure gauge of the production water pump in real time, and transmit the image of the inlet pressure gauge to the hard disk recorder 2 in real time; the outlet side camera 12 is arranged around the outlet of the production water pump and the lens field of view covers the outlet pressure gauge of the production water pump, and is used to collect the image of the outlet pressure gauge of the production water pump in real time, and transmit the image of the outlet pressure gauge to the hard disk recorder 2 in real time. In addition, in order to achieve the purpose of image acquisition day and night, the camera 2 is preferably a camera with infrared night vision function.
[0089] The hard disk recorder 2 is arranged in the target pump station and connected to the wireless router 3 via wired communication, and is used to record the received water inlet pressure gauge image and the water outlet pressure gauge image into the local hard disk in real time, and forward the water inlet pressure gauge image and the water outlet pressure gauge image to the wireless router 3 in real time. Specifically, the hard disk recorder 2 can be, but is not limited to, a Hikvision high-performance hard disk recorder.
[0090] The wireless router 3 is wirelessly connected to the central video host 4, and is used to forward the received inlet pressure gauge image and outlet pressure gauge image to the central video host 4 in real time. Specifically, the wireless router 3 can be, but is not limited to, a 4G router; Figure 1 As shown, one 4G router can be arranged on the target pump station side and the central control side respectively, wherein the 4G router arranged on the target pump station side is connected to the hard disk burner 2 by wired communication, and the 4G router arranged on the central control side is connected to the central video host 4 by wired communication, and the two 4G routers are interconnected through 4G mobile network wireless communication, thereby achieving the purpose of connecting the 4G router arranged on the target pump station side to the central video host 4 by wireless communication.
[0091] The central video host 4 is connected to the central control host 5 by wired communication, and is used to extract the inlet pressure value and the outlet pressure value of the production water pump in real time from the received inlet pressure gauge image and the outlet pressure gauge image, and transmit the inlet pressure value and the outlet pressure value in real time to the central control host 5. The specific extraction means of the inlet pressure value and the outlet pressure value are existing technologies. For example, when the inlet pressure gauge and the outlet pressure gauge of the production water pump respectively display the inlet pressure value and the outlet pressure value through digital display screens, since the inlet pressure gauge image includes a digital display image of the inlet pressure value and the outlet pressure gauge image includes a digital display image of the outlet pressure value, the inlet pressure value and the outlet pressure value can be specifically but not limited to be extracted in real time from the inlet pressure gauge image and the outlet pressure gauge image by conventional OCR (Optical Character Recognition) technology. In addition, the central video host 4 can also be used to output and display the water inlet pressure gauge image and the water outlet pressure gauge image or the water inlet pressure value and the water outlet pressure value in real time.
[0092] The central control host 5 is connected to the radio station transceiver device 6 by wired communication, and is used to generate control instructions for starting the production water pump, stopping the production water pump or adjusting the operating frequency of the production water pump after starting in real time according to the received water inlet pressure value and the water outlet pressure value, and send the control instructions to the radio station transceiver device 6 in real time. The specific generation logic of the aforementioned control instructions may include, but is not limited to: (1) when the water inlet pressure value reaches the target pressure value, if the current state of the production water pump is the start state, an instruction for stopping the production water pump is generated; (2) when the water inlet pressure value is less than the target pressure value, if the current state of the production water pump is the stop state, an instruction for starting the production water pump is generated; (3) when the water inlet pressure value is less than the target pressure value, if the current state of the production water pump is the start state, based on the real-time difference between the water inlet pressure value and the water outlet pressure value, an existing algorithm such as a PID (Proportional-Integral-Derivative control) control algorithm is used to generate a control instruction for adjusting the post-startup operating frequency of the production water pump.
[0093] The radio transceiver device 6 is wirelessly connected to the pump station control cabinet 7, and is used to forward the received control instructions to the pump station control cabinet 7 in real time. Specifically, the radio transceiver device 6 can be, but is not limited to, a long range radio (LoRa) communication device; Figure 1 As shown, a long-distance radio communication device can also be arranged on the target pump station side and the central control side respectively, wherein the long-distance radio communication device arranged on the target pump station side is connected to the pump station control cabinet 7 by wired communication, and the long-distance radio communication device arranged on the central control side is connected to the central control host 5 by wired communication, and the two long-distance radio communication devices are wirelessly connected to each other through LoRa communication technology, thereby achieving the purpose of wirelessly connecting the long-distance radio communication device arranged on the central control side to the pump station control cabinet 7.
[0094] The pump station control cabinet 7 is arranged in the target pump station and is connected to the controlled end of the production water pump by wired communication, and is used to execute the received control instructions in real time to start the production water pump, stop the production water pump, or adjust the operation frequency of the production water pump after starting. Specifically, the pump station control cabinet 7 can be, but is not limited to, a PLC (Programmable Logic Controller) control cabinet of Siemens SMART200 / AE04PLC.
[0095] Therefore, based on the aforementioned production water pump remote control system, a production water pump remote control solution for achieving unattended operation is provided, which includes a camera, a hard disk recorder, a wireless router, a central video host, a central control host, a radio station transceiver and a pump station control cabinet. Through their communication connection relationship and collaboration, the inlet pressure gauge image and the outlet pressure gauge image of the production water pump can be collected in real time, and then the inlet pressure value and the outlet pressure value of the water pump can be extracted in real time from these images. Then, according to these pressure values, control instructions for starting and stopping the water pump or adjusting the operating frequency of the water pump after starting are generated in real time. Finally, the control instructions are wirelessly sent to the pump station control cabinet for execution. In this way, not only can the communication inconvenience between the power plant and the pump station be solved based on wireless communication technology, but there is no need to send people to take turns on duty day and night, and it is conducive to information management, thereby realizing the purpose of automatic control of the pump station, effectively reducing the management and operation costs of the pump station, shortening the maintenance and management response time of the pump station, and improving the management level of the pump station by the enterprise, which is convenient for practical application and promotion.
[0096] Preferably, in order to achieve the purpose of remote monitoring of the production water pump, as Figure 1 As shown, the camera 1 also includes a field monitoring camera 13 connected to the hard disk recorder 2 by wired communication; the field monitoring camera 13 is arranged around the production water pump so that the lens field of view covers the production water pump, and is used to collect the field monitoring image of the production water pump in real time, and transmit the field monitoring image to the hard disk recorder 2 in real time; the hard disk recorder 2 is also used to burn the received field monitoring image into the local hard disk in real time, and forward the field monitoring image to the wireless router 3 in real time; the wireless router 3 is also used to forward the received field monitoring image to the central video host 4 in real time; the central video host 4 is also used to display the received field monitoring image in real time.
[0097] Further preferably, in order to achieve the purpose of identifying water pump faults based on the on-site monitoring images and automatically triggering fault alarms, the central control host 5 is also used to send the control instructions to the central video host 4 in real time; the central video host 4 is also used to, after receiving the control instructions, delay and judge whether the working state of the target pump station matches the control instructions based on the received on-site monitoring images, if not, trigger the execution of the fault alarm action for the target pump station. Since it takes a certain amount of time for the control instructions to be sent to be executed and for the on-site monitoring image to be collected and transmitted to the central video host 4 after execution, it is necessary to delay for a certain period of time (which can be determined based on actual measurements) to judge whether the working state of the target pump station matches the control instructions based on the received on-site monitoring images. In addition, the fault alarm action includes but is not limited to sound and light alarm actions and / or pushing fault alarm messages to management personnel terminals.
[0098] Specifically, if the control instruction is used to start or stop the production water pump, whether the working state of the target pump station is consistent with the control instruction is determined based on the received on-site monitoring image, including but not limited to the following steps S11 to S14.
[0099] S11. Using a target detection algorithm to perform production water pump image recognition processing on the received on-site monitoring image to obtain a water pump image recognition result.
[0100] In the step S11, the target detection algorithm is an existing artificial intelligence recognition algorithm for identifying objects in a picture and marking the location of the objects, and specifically can be but not limited to Faster R-CNN (Faster Regions with Convolutional Neural Networks features, a target detection algorithm proposed by He Kaiming et al. in 2015, which won multiple first places in the 2015 ILSVRV and COCO competitions) target detection algorithm, SSD (Single Shot MultiBox Detector, a target detection algorithm proposed by Wei Liu at ECCV 2016, which is one of the currently popular main detection frameworks) target detection algorithm or YOLO (Youonly look once, and the latest version has been developed to V4, which is also widely used in the industry. Its basic principle is: first, the input image is divided into 7×7 grids, 2 borders are predicted for each grid, and then the target windows with relatively low probability are removed according to the threshold, and finally the redundant windows are removed by border merging to obtain the detection result) target detection algorithm, etc. Therefore, the production water pump image recognition processing can be performed based on the target detection algorithm to obtain the water pump image recognition result.
[0101] S12. According to the production water pump mark box in the water pump image recognition result, an image of the production water pump is captured from the on-site monitoring image.
[0102] S13. Importing the production water pump image into a pre-trained water pump state recognition model based on a convolutional neural network, and outputting a water pump state recognition result.
[0103] In step S13, the convolutional neural network (CNN) is a type of feedforward neural network that includes convolution calculations and has a deep structure. It is one of the representative algorithms of deep learning. It is constructed by imitating the visual perception mechanism of organisms and can perform supervised learning and unsupervised learning. The convolution kernel parameter sharing in the hidden layer and the sparsity of the inter-layer connection enable the convolutional neural network to learn grid-like topology features, such as pixels and audio, with a small amount of calculation, with a stable effect and no additional feature engineering requirements for the data. It is commonly used in computer vision technology. Therefore, based on a certain amount of sample data (i.e., sample images of production water pumps) and label data (i.e., the working state marked for the sample images of production water pumps: start state or stop state), the water pump state recognition model can be obtained by sampling a conventional training method, so that after inputting a test image, the water pump state recognition result of whether the working state is the start state or the stop state and their confidence and other information can be output.
[0104] S14. Determine whether the water pump state identification result matches the control instruction. If so, determine that the working state of the target pump station matches the control instruction. Otherwise, determine that the working state of the target pump station does not match the control instruction.
[0105] In step S14, it is determined whether the state identification result of the production water pump matches the control instruction, specifically including: (1) if the control instruction is used to start the production water pump, it is determined whether the water pump state identification result is a start state, if so, it is determined that the water pump state identification result matches the control instruction, otherwise it does not match; (2) if the control instruction is used to stop the production water pump, it is determined whether the water pump state identification result is a stop state, if so, it is determined that the water pump state identification result matches the control instruction, otherwise it does not match.
[0106] Specifically, if the control instruction is used to adjust the operating frequency of the production water pump after startup, then it is determined whether the working state of the target pump station is consistent with the control instruction based on the received on-site monitoring image, including but not limited to the following steps S21 to S24.
[0107] S21. Use a target detection algorithm to perform production water pump image recognition processing on the received on-site monitoring image to obtain a water pump image recognition result.
[0108] S22. According to the production water pump mark box in the water pump image recognition result, capture the production water pump image from the on-site monitoring image.
[0109] S23. Extracting from the production water pump image the current operating frequency displayed in the working parameter display screen of the production water pump.
[0110] S24. Determine whether the current operating frequency matches the target operating frequency in the control instruction. If so, determine that the working state of the target pump station matches the control instruction. Otherwise, determine that the working state of the target pump station does not match the control instruction.
[0111] The above steps S21-S22 can be derived based on the above steps S11-S12, and will not be repeated here. The specific extraction method of the current operating frequency can also be conventionally implemented by using, but not limited to, OCR technology. In addition, it is determined whether the current operating frequency matches the target operating frequency in the control instruction, specifically but not limited to: determining whether the absolute value of the difference between the current operating frequency and the target operating frequency is less than or equal to a preset threshold value, if so, it is determined to be matched, otherwise it is determined to be mismatched.
[0112] Preferably, considering that the water inlet pressure of the primary pump station often depends on the water level of the water source, and the water level of the water source is greatly affected by external factors, the water inlet pressure of the primary pump station is also directly affected by external factors. Therefore, in order to achieve the purpose of customizing the water pump energy-saving frequency conversion strategy by predicting the water inlet pressure to perform energy-saving control, when the target pump station is a primary pump station, according to the received water inlet pressure value and the water outlet pressure value, a control instruction for starting the production water pump, stopping the production water pump or adjusting the operating frequency of the production water pump after starting is generated in real time, including but not limited to the following steps S101 to S110.
[0113] S101. Acquire water inlet pressure time series data, wherein the water inlet pressure time series data includes but is not limited to a plurality of daily unit time periods that are consecutive in time series and historical water inlet pressure values of each daily unit time period in the plurality of daily unit time periods.
[0114] In the step S101, the intraday unit time period can be, but is not limited to, 1 hour, for example, so that all recorded historical water inlet pressure values can be divided into sub-intervals in the time dimension: starting from the early morning of each day, each day is divided into 24 time periods of 1 hour, and the historical water inlet pressure values in each time period are counted (similar to the stock price, which can be specifically subdivided into the maximum water inlet pressure, the minimum water inlet pressure, the average water inlet pressure or the median water inlet pressure, etc.), and then the statistical results of the historical water inlet pressures of multiple days are combined into the water inlet pressure time series data.
[0115] S102. Divide the water inlet pressure time series data into a plurality of water inlet pressure time series data corresponding one-to-one to a plurality of time windows that are sequentially continuous in time series.
[0116] In step S102, the time window may specifically be, but is not limited to, a natural day, a plurality of consecutive natural days, or a week.
[0117] S103. For the last time window among the multiple time windows, at least one set of water inlet pressure time series data corresponding to at least one previous time window is used as a corresponding training data set, and a set of water inlet pressure time series data corresponding to the corresponding window is used as a corresponding verification data set.
[0118] In step S103, the last time window may specifically be but not limited to this week or this month; the at least one previous time window may specifically be but not limited to last week and the week before last, etc., or the at least one previous time window may specifically be but not limited to last month and the month before last, etc.
[0119] S104. Based on the Prophet model, a real-time dynamic prediction model for water inlet pressure expressed by the following formula is established:
[0120]
[0121] Wherein, t represents time, A(t) represents the water inlet pressure value at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, l q (t) represents the linear regression term at time t and corresponding to the qth linear influence factor among the Q linear influence factors.
[0122] In step S104, the Prophet model is an open source time series prediction tool developed by the Artificial Intelligence Research Institute. It is based on an additive model and can process data with time and seasonal effects. It can be applied to time series prediction problems in various fields, including but not limited to financial market predictions, sales data analysis, and disease incidence monitoring. It provides an easy-to-use and highly flexible framework that allows even non-professional data analysts to effectively perform time series predictions. Therefore, the real-time dynamic prediction model of the water inlet pressure can be obtained based on the regular modification of the Prophet model.
[0123] S105. Set the growth mode of the trend term to a linear growth mode, add natural weeks and natural days to make the seasonal term change periodically, and also set the error noise term to obey a mathematical expectation of μ and a variance of σ 2 The normal distribution of .
[0124] In the step S105, the specific configuration technical details of the trend term, the seasonal term and the error noise term can be conventionally derived by referring to the corresponding term configuration means of the existing Prophet model, and will not be repeated here.
[0125] S106. Obtain the time series data of all water inlet pressure related influencing factors, and then dynamically screen the water inlet pressure linear influencing factors for all the water inlet pressure related influencing factors based on the time series data, so as to determine the water inlet pressure related influencing factors that meet the screening conditions as the linear influencing factors, and obtain a model parameter vector composed of weight adjustment parameters in the linear regression items of all the linear influencing factors.
[0126] In the step S106, the external factors are mainly weather factors, so specifically, all the inlet pressure related influencing factors include but are not limited to wind level, wind speed, time to sunrise time of the day, time to sunset time of the day, cloud cover, perceived temperature, visibility, precipitation probability, humidity, dew point temperature, western Pacific subtropical high intensity index value, India-Myanmar trough intensity index value, Tibet Plateau-1 index value, tropical Indian Ocean sea temperature dipole index value, subtropical South Indian Ocean dipole index value, NINO 3.4 area sea surface temperature anomaly index value and / or sunspot index value (the time series data of these factors can be obtained through conventional online query methods), wherein the day refers to the day corresponding to the unit time period within the day. Specifically, based on the time series data, the linear influencing factors of the inlet pressure are dynamically screened for all the inlet pressure related influencing factors, so as to determine the inlet pressure related influencing factors that meet the screening conditions as the linear influencing factors, including but not limited to the following steps S601 to S605.
[0127] S601. Randomly select historical water inlet pressures of multiple natural days to form a first sample S1, and select data of the multiple natural days from the corresponding time series data as a second sample S2 for a certain water inlet pressure-related influencing factor among all the water inlet pressure-related influencing factors.
[0128] In the step S601, the plurality of natural days may be greater than 60 days, for example.
[0129] S602. Perform a normal distribution KS check on the first sample S1 to calculate a first check statistic p value p1, and also perform the normal distribution KS check on the second sample S2 to calculate a second check statistic p value p2.
[0130] S603. If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the first preset threshold, the correlation coefficient r between the first sample S1 and the sample S2 is calculated according to the following formula: q :
[0131]
[0132] In the formula, i represents a positive integer, x i represents the i-th sample value in the first sample S1, y i represents the i-th sample value in the second sample S2, represents the sample mean of the first sample S1, represents the sample mean of the second sample S2.
[0133] In the step S603, the first preset threshold may be 0.05, for example.
[0134] S604. If the correlation coefficient r q If the inlet pressure is greater than a second preset threshold, it is determined that the certain inlet pressure-related influencing factor meets the screening condition, and the certain inlet pressure-related influencing factor is determined as the linear influencing factor.
[0135] In the step S604, the second preset threshold may be 0.3, for example.
[0136] In the step S106, for the water inlet pressure related influencing factor determined as the linear influencing factor, the corresponding linear regression term l can be specifically recorded as q (t) = r q ×θ q ×M q (t), where θ q Represents the corresponding weight adjustment parameter, M q (t) represents the corresponding numerical value, so that the model parameter vector composed of the weight adjustment parameters in the linear regression terms of all the linear influencing factors can be obtained.
[0137] S107. Apply the training data set and the verification data set corresponding to the last time window, and perform model training and parameter optimization on the real-time dynamic prediction model of the water inlet pressure based on the particle swarm algorithm to obtain a verified model parameter optimization search result, wherein the model parameter optimization search result includes the optimization search result of the model parameter vector, the mathematical expectation and the variance.
[0138] In step S107, the particle swarm optimization (PSO) is a heuristic optimization algorithm that simulates the information sharing and collaboration mechanism in the foraging behavior of bird flocks, and finds the optimal solution to the problem through interaction between individuals. In order to balance the local optimum and global optimum capabilities in parameter optimization, improve the efficiency of optimizing parameter search while ensuring overall reliability, preferably, the training data set and the verification data set corresponding to the last time window are used to perform model training and parameter optimization on the water inlet pressure real-time dynamic prediction model based on the particle swarm algorithm, and obtain the verified model parameter optimization search results, including but not limited to the following steps S701 to S712.
[0139] S701. Initialize the algorithm parameters of the particle swarm algorithm including the population size N, the maximum number of iterations T, the self-learning factor C1 and the global learning factor C2, and randomly generate the initial position of each particle in the particle population, and then execute step S702, wherein the initial position of each particle is randomly generated based on the following formula:
[0140]
[0141] In the formula, i' represents a positive integer, D represents a positive integer less than or equal to D, D represents the number of dimensions of the model parameter vector to be optimized, and the model parameter vector to be optimized includes the model parameter vector to be optimized, the Mathematical Expectation and stated variance , represents the initial position of the i'th particle in the particle population on the dth dimension of the model parameter vector to be optimized, u c,d represents the upper limit of the model parameter search space in the dth dimension, l c,d represents the lower limit of the model parameter search space in the d-th dimension, and rand(0,1) represents a pure decimal random generation function.
[0142] S702. For each particle, the corresponding initial position is input into the water inlet pressure real-time dynamic prediction model as the model parameter search result, and then the training data set corresponding to the last time window is used to train the water inlet pressure real-time dynamic prediction model, and the corresponding fitness F is calculated, and then step S703 is executed, wherein the fitness F is calculated according to the following formula:
[0143]
[0144] Where n represents the total number of samples in the training data set, i represents a positive integer, represents the prediction output item calculated after the model input item of the i-th sample data in the training data set is imported into the water inlet pressure real-time dynamic prediction model, y (i) Represents the true output item of the i-th sample data.
[0145] In step S702, the specific process of the model training is the existing technology.
[0146] S703. Take the initial position of a particle with the minimum fitness as the initial global optimal position G ini , the position of the i'th particle in the iteration process and corresponding to the minimum fitness is the local optimal position P reached by the i'th particle i’ , and the position of the particle population in the iteration process and corresponding to the minimum fitness is recorded as the global optimal position G reached by the particle population, and the moving speed of the i'th particle is recorded as V i’ , and finally initialize the local optimal position P i’ =G=G ini , and initialize the moving speed V i’ is zero, and the particle movement inertia coefficient range is also set to [T min ,T max ], and initialize the period T of the periodic reduction of the inertia coefficient to a randomly selected value within the range of the particle movement inertia coefficient, and also initialize the particle movement inertia coefficient during the iteration process to a value greater than zero, and also initialize the current iteration number t' to 0, and then execute step S704.
[0147] S704. Update the particle movement inertia coefficient of each particle, and then execute step S705, wherein the update formula of the particle movement inertia coefficient is as follows:
[0148]
[0149] In the formula, represents the particle movement inertia coefficient of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the particle movement inertia coefficient of the i'th particle in the dth dimension before the t'+1th iteration, t'%T represents the modulo operation result of the previous iteration number t' on the period T, and if If it is less than 0, a new value is randomly selected within the range of the particle's inertia coefficient as the new period T, and the calculation is recalculated. until Not less than 0.
[0150] S705. Update the moving speed of each particle, and then execute step S706, wherein the updating formula of the moving speed is as follows:
[0151]
[0152] In the formula, represents the moving speed of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the moving speed of the i'th particle in the dth dimension before the t'+1th iteration, γ1 represents the first random number uniformly distributed in (0,1), represents the local optimal position reached by the i'th particle in the dth dimension during the first t' iterations, represents the position of the i'th particle in the dth dimension before the t'+1th iteration, γ2 represents a second random number uniformly distributed in (0,1), represents the global optimal position reached by the particle population in the d-th dimension during the first t' iterations.
[0153] S706. Update the position of each particle, and then execute step S707, wherein the update formula of the position is as follows:
[0154] In the formula, represents the position of the i'th particle in the dth dimension obtained after the t'+1th iteration.
[0155] S707. For each particle, the corresponding position is input into the real-time dynamic prediction model of water inlet pressure as the search result of the model parameter, and then the training data set corresponding to the last time window is applied again to train the real-time dynamic prediction model of water inlet pressure, and the corresponding fitness F is calculated again, and then step S708 is executed.
[0156] S708. Determine the local optimal position P i’ Is the corresponding fitness greater than the current fitness of the i'th particle? If so, the local optimal position P i’ Update to the current position of the i'th particle, and then execute step S709; otherwise, directly execute step S709.
[0157] S709. Determine whether the fitness corresponding to the global optimal position G is greater than the minimum value among the current fitness of each particle. If so, update the global optimal position G to the current position of any particle with the minimum value, and then execute step S710; otherwise, directly execute step S710.
[0158] S710. Increment the current number of iterations t' by 1, and determine whether the current number of iterations t' reaches the maximum number of iterations T. If so, execute step S711; otherwise, return to execute step S704.
[0159] S711. Input the global optimal position G as the model parameter search result into the water inlet pressure real-time dynamic prediction model, and then use the water inlet pressure real-time dynamic prediction model and the verification data set corresponding to the last time window to predict the water inlet pressure, and calculate the new fitness F new , then execute step S712;
[0160] S712. Judge F new <ρ×F min Is it true? If so, the global optimal position G is used as the verified model parameter optimization search result. Otherwise, after adjusting the search space of the model parameter vector to be optimized, return to step S701, where ρ represents a preset error verification coefficient and is greater than 1, and F min represents the fitness corresponding to the global optimal position G, and the model parameter optimization search result includes the optimization search results of the model parameter vector, the mathematical expectation and the variance.
[0161] S108. Input the search results of the model parameter optimization into the real-time dynamic prediction model for water inlet pressure, and then apply the real-time dynamic prediction model for water inlet pressure to predict the predicted value of water inlet pressure in a unit time period within the current day, a unit time period within the next day, or a unit time period within at least two days in the future.
[0162] In the step S108, for example, if the current intraday unit time period is 12:00 to 13:00, then the current next intraday unit time period is 13:00 to 14:00; the current at least two future intraday unit time periods include 13:00 to 14:00, 14:00 to 15:00 and 15:00 to 16:00, etc.
[0163] S109. According to a plurality of water inlet pressure value ranges that are preset and correspond one-to-one to a plurality of water pump energy-saving frequency conversion strategies, a certain water inlet pressure value range that includes the water inlet pressure prediction value is determined from the plurality of water inlet pressure value ranges.
[0164] In step S109, the water pump energy-saving frequency conversion strategy may specifically include but is not limited to: a water pump operating frequency upper limit value, a water pump operating frequency lower limit value and / or a water pump operating frequency change step size preset for energy saving purposes.
[0165] S110. According to a water pump energy-saving and frequency-conversion strategy among the multiple water pump energy-saving and frequency-conversion strategies and corresponding to the certain water inlet pressure value range, a control instruction for starting the production water pump, stopping the production water pump or adjusting the operating frequency of the production water pump after starting is generated in real time according to the received water inlet pressure value and the water outlet pressure value.
[0166] In summary, the production water pump remote control system provided by this embodiment has the following technical effects:
[0167] (1) This embodiment provides a remote control solution for production water pumps for achieving unattended operation, which includes a camera, a hard disk recorder, a wireless router, a central video host, a central control host, a radio station transceiver and a pump station control cabinet. Through their communication connection relationship and cooperation, the images of the water inlet pressure gauge and the water outlet pressure gauge of the production water pump can be collected in real time, and then the water inlet pressure value and the water outlet pressure value of the water pump can be extracted in real time from these images. Then, according to these pressure values, control instructions for starting and stopping the water pump or adjusting the operating frequency of the water pump after starting are generated in real time. Finally, the control instructions are wirelessly sent to the pump station control cabinet for execution. In this way, not only can the communication inconvenience between the power plant and the pump station be solved based on wireless communication technology, but there is no need to send people to take turns on duty day and night, and it is conducive to information management, thereby realizing the purpose of automatic control of the pump station, effectively reducing the management and operation costs of the pump station, shortening the response time of the pump station maintenance management, and improving the management level of the pump station by the enterprise, which is convenient for practical application and promotion.
[0168] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A remote control system for a production water pump, characterized in that: It includes a camera (1), a hard disk recorder (2), a wireless router (3), a central video host (4), a central control host (5), a radio station transceiver (6) and a pump station control cabinet (7); The camera (1) is arranged in the target pump station and connected to the hard disk recorder (2) by wired communication, and is used to collect images of the water inlet pressure gauge and the water outlet pressure gauge of the production water pump in the target pump station in real time, and transmit the images of the water inlet pressure gauge and the water outlet pressure gauge to the hard disk recorder (2) in real time; The hard disk recorder (2) is arranged in the target pump station and connected to the wireless router (3) via wired communication, and is used to record the received water inlet pressure gauge image and the water outlet pressure gauge image into a local hard disk in real time, and forward the water inlet pressure gauge image and the water outlet pressure gauge image to the wireless router (3) in real time; The wireless router (3) is wirelessly connected to the central video host (4) and is used to forward the received water inlet pressure gauge image and the water outlet pressure gauge image to the central video host (4) in real time; The central video host (4) is connected to the central control host (5) by wired communication, and is used to extract the water inlet pressure value and the water outlet pressure value of the production water pump in real time from the received water inlet pressure gauge image and the water outlet pressure gauge image, and transmit the water inlet pressure value and the water outlet pressure value to the central control host (5) in real time; The central control host (5) is connected to the radio station transceiver (6) by wired communication, and is used to generate a control instruction for starting the production water pump, stopping the production water pump, or adjusting the operating frequency of the production water pump after starting according to the received water inlet pressure value and the water outlet pressure value, and send the control instruction to the radio station transceiver (6) in real time; The radio station transceiver device (6) is wirelessly connected to the pump station control cabinet (7) and is used to forward the received control instructions to the pump station control cabinet (7) in real time; The pump station control cabinet (7) is arranged in the target pump station and is connected to the controlled end of the production water pump by wired communication, and is used to execute the received control instructions in real time to start the production water pump, stop the production water pump or adjust the operating frequency of the production water pump after starting.
2. The production water pump remote control system according to claim 1, characterized in that: The camera (1) comprises a water inlet side camera (11) and a water outlet side camera (12) which are respectively connected to the hard disk recorder (2) by wired communication; The water inlet side camera (11) is arranged around the water inlet of the production water pump and has a lens field of view covering the water inlet pressure gauge of the production water pump, and is used to collect an image of the water inlet pressure gauge of the production water pump in real time, and transmit the image of the water inlet pressure gauge to the hard disk recorder (2) in real time; The water outlet side camera (12) is arranged around the water outlet of the production water pump so that the lens field of view covers the water outlet pressure gauge of the production water pump, and is used to collect the image of the water outlet pressure gauge of the production water pump in real time, and transmit the image of the water outlet pressure gauge to the hard disk recorder (2) in real time.
3. The production water pump remote control system according to claim 2, characterized in that: The camera (1) further comprises a field monitoring camera (13) connected to the hard disk recorder (2) via wired communication; The on-site monitoring camera (13) is arranged around the production water pump so that the camera field of view covers the production water pump, and is used to collect on-site monitoring images of the production water pump in real time, and transmit the on-site monitoring images to the hard disk recorder (2) in real time; The hard disk recorder (2) is also used to burn the received on-site monitoring images into a local hard disk in real time, and forward the on-site monitoring images to the wireless router (3) in real time; The wireless router (3) is also used to forward the received on-site monitoring images to the central video host (4) in real time; The central video host (4) is also used to display the received on-site monitoring images in real time.
4. The production water pump remote control system according to claim 3, characterized in that: The central control host (5) is also used to send the control instruction to the central video host (4) in real time; The central video host (4) is also used to, after receiving the control instruction, delay and determine whether the working status of the target pump station is consistent with the control instruction based on the received on-site monitoring image, and if not, trigger the execution of a fault alarm action for the target pump station.
5. The production water pump remote control system according to claim 4, characterized in that: If the control instruction is used to start the production water pump or stop the production water pump, judging whether the working state of the target pump station matches the control instruction according to the received on-site monitoring image includes: Using a target detection algorithm to perform production water pump image recognition processing on the received on-site monitoring image to obtain a water pump image recognition result; According to the production water pump mark frame in the water pump image recognition result, an image of the production water pump is intercepted from the on-site monitoring image; Importing the production water pump image into a water pump state recognition model pre-trained based on a convolutional neural network, and outputting a water pump state recognition result; It is determined whether the water pump state identification result matches the control instruction. If so, it is determined that the working state of the target pump station matches the control instruction. Otherwise, it is determined that the working state of the target pump station does not match the control instruction.
6. The production water pump remote control system according to claim 4, characterized in that: If the control instruction is used to adjust the operating frequency of the production water pump after startup, judging whether the working state of the target pump station matches the control instruction according to the received on-site monitoring image includes: Using a target detection algorithm to perform production water pump image recognition processing on the received on-site monitoring image to obtain a water pump image recognition result; According to the production water pump mark frame in the water pump image recognition result, an image of the production water pump is intercepted from the on-site monitoring image; Extracting from the production water pump image a current operating frequency displayed in a working parameter display screen of the production water pump; It is determined whether the current operating frequency matches the target operating frequency in the control instruction. If so, it is determined that the working state of the target pump station matches the control instruction. Otherwise, it is determined that the working state of the target pump station does not match the control instruction.
7. The production water pump remote control system according to claim 1, characterized in that: When the target pump station is a primary pump station, a control instruction for starting the production water pump, stopping the production water pump, or adjusting the operation frequency of the production water pump after starting is generated in real time according to the received water inlet pressure value and the water outlet pressure value, including: Acquire water inlet pressure time series data, wherein the water inlet pressure time series data contains a plurality of intra-day unit time periods that are sequentially continuous in time series and a historical water inlet pressure value of each intra-day unit time period in the plurality of intra-day unit time periods; Dividing the water inlet pressure time series data into a plurality of water inlet pressure time series data corresponding one to one to a plurality of time windows that are sequentially continuous in time series; For the last time window among the multiple time windows, at least one piece of water inlet pressure time series data corresponding to at least one corresponding previous time window is used as a corresponding training data set, and a piece of water inlet pressure time series data corresponding to the corresponding window is used as a corresponding verification data set; Based on the Prophet model, a real-time dynamic prediction model for water inlet pressure is established using the following formula: Wherein, t represents time, A(t) represents the water inlet pressure value at time t, g(t) represents the trend term at time t, s(t) represents the seasonal term at time t, ε(t) represents the error noise term at time t, Q represents a positive integer, q represents a positive integer less than or equal to Q, l q (t) represents the linear regression term corresponding to the qth linear influence factor among the Q linear influence factors at time t; The growth mode of the trend term is set to a linear growth mode, and natural weeks and natural days are added to make the seasonal term change periodically, and the error noise term is also set to obey a mathematical expectation of μ and a variance of σ 2 Normal distribution of Acquire the time series data of all water inlet pressure related influencing factors, and then dynamically screen the water inlet pressure linear influencing factors for all the water inlet pressure related influencing factors based on the time series data, so as to determine the water inlet pressure related influencing factors that meet the screening conditions as the linear influencing factors, and obtain a model parameter vector composed of weight adjustment parameters in the linear regression items of all the linear influencing factors; Using the training data set and the verification data set corresponding to the last time window, the real-time dynamic prediction model of the water inlet pressure is trained and optimized based on the particle swarm algorithm to obtain a verified model parameter optimization search result, wherein the model parameter optimization search result includes the optimization search result of the model parameter vector, the mathematical expectation and the variance; Inputting the search results of the model parameter optimization into the real-time dynamic prediction model for water inlet pressure, and then applying the real-time dynamic prediction model for water inlet pressure to predict the predicted value of water inlet pressure in a unit time period within the current day, a unit time period within the next day, or a unit time period within at least two days in the future; According to a plurality of water inlet pressure value ranges that are preset and correspond one to one with a plurality of water pump energy-saving frequency conversion strategies, a certain water inlet pressure value range that includes the water inlet pressure prediction value is determined from the plurality of water inlet pressure value ranges; According to a water pump energy-saving and frequency-changing strategy among the multiple water pump energy-saving and frequency-changing strategies and corresponding to the certain water inlet pressure value range, a control instruction for starting the production water pump, stopping the production water pump or adjusting the operating frequency of the production water pump after starting is generated in real time according to the received water inlet pressure value and the water outlet pressure value.
8. The production water pump remote control system according to claim 7, characterized in that: Dynamically screening the linear influencing factors of the water inlet pressure for all the inlet pressure-related influencing factors based on the time series data, so as to determine the inlet pressure-related influencing factors that meet the screening conditions as the linear influencing factors, includes: Randomly select historical water inlet pressures of multiple natural days to form a first sample S1, and select data of the multiple natural days from the corresponding time series data as a second sample S2 for a certain water inlet pressure-related influencing factor among all the water inlet pressure-related influencing factors; Performing a normal distribution KS check on the first sample S1 to calculate a first check statistic p value p1, and also performing the normal distribution KS check on the second sample S2 to calculate a second check statistic p value p2; If the first check statistic p value p1 and the second check statistic p value p2 are both greater than the first preset threshold, the correlation coefficient r between the first sample S1 and the sample S2 is calculated according to the following formula: q : In the formula, i represents a positive integer, x i represents the i-th sample value in the first sample S1, y i represents the i-th sample value in the second sample S2, represents the sample mean of the first sample S1, represents the sample mean of the second sample S2; If the correlation coefficient r q If the inlet pressure is greater than a second preset threshold, it is determined that the certain inlet pressure-related influencing factor meets the screening condition, and the certain inlet pressure-related influencing factor is determined as the linear influencing factor.
9. The production water pump remote control system according to claim 7, characterized in that: Using the training data set and the verification data set corresponding to the last time window, the real-time dynamic prediction model of the water inlet pressure is trained and optimized based on the particle swarm algorithm to obtain the verified model parameter optimization search result, including the following steps S701 to S712: S701. Initialize the algorithm parameters of the particle swarm algorithm including the population size N, the maximum number of iterations T, the self-learning factor C1 and the global learning factor C2, and randomly generate the initial position of each particle in the particle population, and then execute step S702, wherein the initial position of each particle is randomly generated based on the following formula: Wherein, i' represents a positive integer, d represents a positive integer less than or equal to D, D represents the number of dimensions of the model parameter vector to be optimized, and the model parameter vector to be optimized includes the model parameter vector to be optimized, the mathematical expectation and the variance. represents the initial position of the i'th particle in the particle population on the dth dimension of the model parameter vector to be optimized, u c,d represents the upper limit of the model parameter search space in the dth dimension, l c,d represents the lower limit of the model parameter search space in the d-th dimension, and rand(0,1) represents a pure decimal random generator function; S702. For each particle, the corresponding initial position is input into the water inlet pressure real-time dynamic prediction model as the model parameter search result, and then the training data set corresponding to the last time window is used to train the water inlet pressure real-time dynamic prediction model, and the corresponding fitness F is calculated, and then step S703 is executed, wherein the fitness F is calculated according to the following formula: Where n represents the total number of samples in the training data set, i represents a positive integer, represents the prediction output item calculated after the model input item of the i-th sample data in the training data set is imported into the water inlet pressure real-time dynamic prediction model, y (i) represents the true output item of the i-th sample data; S703. Take the initial position of a particle with the minimum fitness as the initial global optimal position G ini , the position of the i'th particle in the iteration process and corresponding to the minimum fitness is the local optimal position P reached by the i'th particle i’ , and the position of the particle population in the iteration process and corresponding to the minimum fitness is recorded as the global optimal position G reached by the particle population, and the moving speed of the i'th particle is recorded as V i’ Finally, initialize and set the local optimal position P i’ =G=G ini , and initialize the moving speed V i’ is zero, and the particle movement inertia coefficient range is also set to [T min ,T max ], and initialize the period T of the periodic reduction of the inertia coefficient to a randomly selected value within the range of the particle movement inertia coefficient, and also initialize the particle movement inertia coefficient in the iteration process to a value greater than zero, and also initialize the current iteration number t' to 0, and then execute step S704; S704. Update the particle movement inertia coefficient of each particle, and then execute step S705, wherein the update formula of the particle movement inertia coefficient is as follows: In the formula, represents the particle movement inertia coefficient of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the particle movement inertia coefficient of the i'th particle in the dth dimension before the t'+1th iteration, t'%T represents the modulo operation result of the previous iteration number t' on the period T, and if If it is less than 0, a new value is randomly selected within the range of the particle's inertia coefficient as the new period T, and the calculation is recalculated. until Not less than 0; S705. Update the moving speed of each particle, and then execute step S706, wherein the updating formula of the moving speed is as follows: In the formula, represents the moving speed of the i'th particle in the dth dimension obtained after the t'+1th iteration, represents the moving speed of the i'th particle in the dth dimension before the t'+1th iteration, γ1 represents the first random number uniformly distributed in (0,1), represents the local optimal position reached by the i'th particle in the dth dimension during the first t' iterations, represents the position of the i'th particle in the dth dimension before the t'+1th iteration, β2 represents a second random number uniformly distributed in (0,1), represents the global optimal position reached by the particle population in the d-th dimension during the first t' iterations; S706. Update the position of each particle, and then execute step S707, wherein the update formula of the position is as follows: In the formula, represents the position of the i'th particle in the dth dimension obtained after the t'+1th iteration; S707. For each particle, the corresponding position is input into the water inlet pressure real-time dynamic prediction model as the model parameter search result, and then the training data set corresponding to the last time window is used again to perform model training on the water inlet pressure real-time dynamic prediction model, and the corresponding fitness F is calculated again, and then step S708 is executed; S708. Determine the local optimal position P i’ Is the corresponding fitness greater than the current fitness of the i'th particle? If so, the local optimal position P i’ Update to the current position of the i'th particle, and then execute step S709, otherwise directly execute step S709; S709. Determine whether the fitness corresponding to the global optimal position G is greater than the minimum value among the current fitness of each particle. If so, update the global optimal position G to the current position of any particle with the minimum value, and then execute step S710. Otherwise, directly execute step S710; S710. Increment the current number of iterations t' by 1, and determine whether the current number of iterations t' reaches the maximum number of iterations T. If so, execute step S711, otherwise return to execute step S704; S711. Input the global optimal position G as the model parameter search result into the water inlet pressure real-time dynamic prediction model, and then use the water inlet pressure real-time dynamic prediction model and the verification data set corresponding to the last time window to predict the water inlet pressure, and calculate the new fitness F new , then execute step S712; S712. Judge F new <ρ×F min Is it true? If so, the global optimal position G is used as the verified model parameter optimization search result. Otherwise, after adjusting the search space of the model parameter vector to be optimized, return to step S701, where ρ represents a preset error verification coefficient and is greater than 1, and F min represents the fitness corresponding to the global optimal position G, and the model parameter optimization search result includes the optimization search results of the model parameter vector, the mathematical expectation and the variance.
10. The production water pump remote control system according to claim 7, characterized in that: All water inlet pressure-related influencing factors include wind level, wind speed, time to sunrise on the same day, time to sunset on the same day, cloud cover, perceived temperature, visibility, precipitation probability, humidity, dew point temperature, western Pacific subtropical high intensity index value, Indo-Myanmar trough intensity index value, Tibet Plateau-1 index value, tropical Indian Ocean sea surface temperature dipole index value, subtropical South Indian Ocean dipole index value, NINO 3.4 region sea surface temperature anomaly index value and / or sunspot index value, wherein the current day refers to the day corresponding to the unit time period within the day.
Citation Information
Patent Citations
Automatic relay water supply equipment control system for super high-rise and control method
CN101220605A
Non-negative-pressure pump station centralized remote monitoring system
CN107561995A
Draining pump station optimization scheduling method based on machine learning technology
CN114792071A
Water pump unit equipment energy-saving operation system based on Internet of Things
CN115853759A
Unattended pressurization station control system based on AI algorithm
CN116699998A
Cited By
Remote control method, device, system, equipment and product for power converter
CN119154642A
