Distributed pump set anti-wear scheduling method and system, electronic device, and storage medium

Abstract

The application discloses a distributed pump group anti-bending wear scheduling method and system, electronic equipment and a storage medium, and belongs to the pump group control technical field of a fluid conveying system. The method comprises the following steps: determining the number of segments and the maximum number of pumps in a single segment, and constructing a pump group state matrix; acquiring real-time monitoring data of each segment in the pump group state matrix; when the real-time monitoring data of the segment meets a preset triggering condition, a scheduling triggering signal of the segment is generated; based on the scheduling triggering signal, a state vector of the segment is extracted to construct a decision vector of the segment; based on a preset rule, a pump index corresponding to a target component is selected from the decision vector, and a pump to be scheduled is determined based on the pump index.

Description

Technical Field

[0001] This application relates to the field of pump control technology for fluid transport systems, specifically to a method, system, electronic equipment, and storage medium for anti-wear scheduling of distributed pump sets. Background Technology

[0002] Current mainstream scheduling methods in the industry have significant drawbacks. On the one hand, the fixed-sequence pump start-up strategy is simple in logic and has a fast response, but running in a fixed order for a long time will lead to excessive wear on some equipment (i.e., "uneven wear"), causing the difference in the cumulative running time of core equipment to increase linearly or even exponentially, significantly shortening the overall service life of the pump group and increasing the cost of equipment replacement and maintenance. On the other hand, intelligent algorithm scheduling schemes rely on a large amount of operating data for model training, which has a long data collection cycle and high cost. Moreover, the generalization ability of the model is often limited by the quality and range of training data, and its adaptability to parameter changes in different scenarios is poor.

[0003] Further analysis reveals that the number of pump groups in different application scenarios and the number of pump groups in a single segment vary significantly. Existing technologies lack a standardized scheduling scheme that can overcome scenario and configuration limitations while balancing theoretical rigor and engineering practicality. Summary of the Invention

[0004] This invention provides a method, system, electronic equipment, and storage medium for scheduling distributed pump sets to prevent uneven wear, aiming to solve common problems such as uneven wear loss, configuration limitations, and complex deployment in multi-device clusters.

[0005] Firstly, a method for scheduling distributed pump sets to prevent uneven wear is provided, comprising the following steps: Determine the number of segments and the maximum number of pumps per segment, and construct the pump group state matrix; The real-time monitoring data of each segment in the pump group status matrix is ​​obtained, and when the real-time monitoring data of the segment meets the preset triggering conditions, the scheduling trigger signal of the segment is generated. Based on the scheduling trigger signal, the state vector of the segment is extracted to construct the decision vector of the segment; Based on preset rules, the pump index corresponding to the target component is selected from the decision vector, and the pump to be scheduled is determined based on the pump index.

[0006] In some design approaches, extracting the state vector of the segment to construct the decision vector of the segment includes: Based on the state vector of the segment, a selection vector for the available pumps of the segment is constructed; The decision vector is calculated based on the filtering vector and combined with the random number vector.

[0007] In some of these designs, the random number vector is generated based on a hardware random number generator, a Mason tween algorithm, or a linear congruence generator.

[0008] In some of these design approaches, the selection vector for the available pumps used to construct the segment also includes: If the number of available pumps in the segment is zero, a fault alarm is triggered and the current scheduling is terminated.

[0009] In some of these design approaches, after constructing the pump set state matrix, the method further includes: The pump group state matrix is ​​initialized based on the encoded values ​​corresponding to the initial states of all pumps. The initial state of the pump includes standby state, running state, and fault state.

[0010] In some of these design approaches, after constructing the pump set state matrix, the method further includes: If the actual number of pumps in a segment is less than the maximum number of pumps in a single segment, the element in the pump group state matrix corresponding to the segment without a pump is initialized to a missing code value.

[0011] In some of these design approaches, the real-time monitoring data for each segment includes a pressure value and the duration of that pressure value; the preset triggering conditions include a first pressure threshold and a first duration threshold. The generation of the segmented scheduling trigger signal includes: The pressure value of the segment and the duration of the pressure value are compared with the corresponding first pressure threshold and first duration threshold, respectively. If the pressure value of the segment is less than the first pressure threshold and its duration is greater than the first duration threshold, a scheduling trigger signal is generated.

[0012] In some of these design approaches, the preset rule is the maximum component rule; The step of selecting the pump index corresponding to the target component from the decision vector based on preset rules includes: The largest component is selected from the decision vector as the target component, and the pump to be scheduled is determined based on its corresponding pump index.

[0013] In some of these design approaches, the preset rule is the minimum component rule; The step of selecting the pump index corresponding to the target component from the decision vector based on preset rules includes: The minimum component is selected from the decision vector as the target component, and the pump to be scheduled is determined based on its corresponding pump index.

[0014] In some design approaches, the step of selecting the pump index corresponding to the target component from the decision vector based on preset rules further includes: If there are multiple identical components in the decision vector, the component with the smallest pump index is selected as the target component.

[0015] In some of these design approaches, after determining each pump to be scheduled, the method further includes: Start each of the scheduled pumps and record the start timestamp; The real-time monitoring data of each segment containing the pump to be scheduled is continuously monitored. If the preset stop conditions are met, the pump with the longest running time is shut down based on the current timestamp.

[0016] In some of these design approaches, the segmented real-time monitoring data includes a pressure value and the duration of that pressure value; the preset stop condition includes a second pressure threshold and a second duration threshold. Determining whether the preset stop condition is met includes: The pressure value of the segment and the duration of the pressure value are compared with the corresponding second pressure threshold and second duration threshold, respectively. If the pressure value of the segment is greater than or equal to the second pressure threshold and its duration is greater than the second duration threshold, then the preset stop condition is satisfied.

[0017] In some of these designs, after starting / stopping each of the scheduled pumps, the method further includes: The status of the pump to be scheduled in the pump group status matrix is ​​updated accordingly.

[0018] Secondly, a distributed pump set anti-wear scheduling system is also provided, which applies the distributed pump set anti-wear scheduling method described above. The scheduling system includes: The matrix construction module is used to determine the number of segments and the maximum number of pumps per segment, construct the pump group state matrix, and initialize the pump group state matrix. The signal generation module is used to acquire real-time monitoring data of each segment in the pump group status matrix, and generate a scheduling trigger signal for the segment when the real-time monitoring data of the segment meets the preset trigger conditions. The vector calculation module is used to extract the state vector of the segment based on the scheduling trigger signal, so as to construct the decision vector of the segment; The scheduling selection module is used to select the pump index corresponding to the target component from the decision vector based on preset rules, and determine the pump to be scheduled based on the pump index.

[0019] Thirdly, an electronic device is also provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, it implements the distributed pump group anti-wear scheduling method as described above.

[0020] Fourthly, a computer-readable storage medium is also provided, on which a computer program is stored, the computer program being loaded by a processor to perform the steps in the distributed pump set anti-wear scheduling method described above.

[0021] Beneficial effects: The distributed pump set anti-wear scheduling method of this application, by introducing a state matrix and a missing state isolation mechanism, realizes standardized and unified management of any number of segments and any number of pumps in a single segment, breaking the limitations of scenario and physical configuration; at the same time, it formulates scheduling strategies based on real-time monitoring data of segments, ensuring that all available pumps are called with equal probability in the long term, fundamentally eliminating the problem of uneven wear, extending the overall service life of the equipment, and solving the core problems of poor universality, easy uneven wear, and difficult engineering deployment of traditional pump set scheduling. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart illustrating the distributed pump set anti-wear scheduling method provided in an exemplary embodiment of this application; Figure 2 This is another flowchart illustrating the distributed pump set anti-wear scheduling method provided in an exemplary embodiment of this application; Figure 3 This is a comparison chart of the standard deviation between the distributed pump set anti-wear scheduling method provided by the exemplary embodiments of this application and the traditional FIFO scheduling method; Figure 4 This is a comparison diagram of the relative deviation between the distributed pump set anti-wear scheduling method provided in the exemplary embodiments of this application and the traditional FIFO scheduling method; Figure 5 This is a schematic diagram of the modules of the distributed pump set anti-wear scheduling system provided in an exemplary embodiment of this application; Figure 6 This is a schematic diagram illustrating the application principle of the distributed pump set anti-wear scheduling system provided in an exemplary embodiment of this application.

[0024] In the diagram: 10, matrix construction module; 20, signal generation module; 30, vector calculation module; 40, scheduling selection module. Detailed Implementation

[0025] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0026] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0027] "A and / or B" includes the following three combinations: A only, B only, and a combination of A and B.

[0028] The use of "applies to" or "configured to" in this application implies open and inclusive language, which does not exclude the applicability to or configuration to devices performing additional tasks or steps. Additionally, the use of "based on" implies openness and inclusivity, because processes, steps, calculations, or other actions "based on" one or more of the stated conditions or values ​​may in practice be based on additional conditions or values ​​beyond those stated.

[0029] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0030] As the core power unit of fluid transport systems, pump sets are commonly used in segmented or distributed layouts in scenarios such as ship fire fighting and industrial water supply. Their scheduling efficiency directly affects the system's operational stability and equipment lifespan. According to existing technologies, the number of pump sets in different application scenarios and the number of pump sets within a single segment vary significantly, lacking a standardized scheduling scheme that can overcome scenario and configuration limitations while balancing theoretical rigor and engineering practicality. Extending to the field of multi-device clusters, traditional scheduling of multiple generator sets, charging piles, and other equipment also faces common problems such as uneven losses due to fixed sequences and high barriers to deploying intelligent algorithms. Furthermore, existing scheduling algorithms are highly complex, requiring substantial hardware resources for computation, making efficient deployment difficult in low-configuration industrial hardware such as shipboard PLCs and embedded controllers.

[0031] Based on this, this application provides a distributed pump set anti-wear scheduling method, focusing on the balanced scheduling and anti-wear control of segmented / distributed pump sets, and is deeply adapted to multiple scenarios such as marine fire protection systems, industrial water supply systems, building booster systems, and chemical fluid transportation systems. Its core advantage lies in breaking through the limitations of traditional technical configurations, supporting flexible combinations of any number of segments and pumps per segment, and achieving long-term stable and load-balanced operation of the pump set.

[0032] Firstly, this embodiment provides a distributed pump set anti-wear scheduling method; please refer to [link to relevant documentation]. Figure 1 and Figure 2 This includes the following steps: S100. Determine the number of segments and the maximum number of pumps per segment, and construct the pump group state matrix.

[0033] Specifically, firstly, based on the actual system topology, the number of segments M and the maximum number of pumps per segment K are determined, where M and K are both positive integers. Understandably, in practical applications, this can be extended to any number of segments and any number of pump groups per segment, depending on the scenario. Then, based on the determined number of segments M and the maximum number of pumps per segment K, an M×K dimensional pump group state matrix Q is constructed.

[0034] Subsequently, the encoded values ​​corresponding to the initial states of all pumps are read to initialize the constructed pump group state matrix Q. Generally, pump states include standby, running, and fault states. In some implementations, the standby state can be encoded as 0, the running state as 1, and the fault state as 2. Further, in some embodiments, for cases where the actual number of pumps in a segment is less than the maximum number of pumps in a single segment, the elements in the pump group state matrix corresponding to no pumps in that segment can be initialized to empty encoded values. Specifically, assuming the actual number of pumps in a segment n (n is a positive integer) is less than the maximum number of pumps in a single segment K (i.e., n < K), then the elements in the pump group state matrix Q corresponding to the (n+1)th to the Kth column of that segment can be uniformly initialized to empty encoded values. In this embodiment, the empty encoded value can be set to 3, indicating an empty / no physical pump.

[0035] Therefore, the pump state matrix Q can be initialized and configured based on the encoded values ​​corresponding to the initial states of all pumps. A two-way verification mechanism with the physical control system or monitoring system can be used to ensure that the matrix state is consistent with the physical system. In essence, each row of the pump state matrix Q represents a segment, and each column represents the initial state of the pump in the corresponding segment.

[0036] Thus, by assigning different coding values ​​to different pump states and independently coding vacant states, the state matrix dimension M×K can be flexibly adjusted. The isolation mechanism between vacant positions (state code 3) and actual equipment states (codes 0, 1, 2) breaks the coupling limitation between the number of segments and the number of pumps per segment, enabling the manipulation of any M (positive integer) and... The standardized management of (positive integer) configuration scenarios has the ability to adapt to various scenarios, breaking the limitations of scenarios and physical configurations.

[0037] S200: Obtain real-time monitoring data of each segment in the pump group status matrix. When the real-time monitoring data of a segment meets the preset triggering conditions, generate a segment scheduling trigger signal.

[0038] In some implementations, real-time monitoring data for each segment of the pump set state matrix can be acquired by setting sensors. Specifically, each segment of the pump set state matrix can be equipped with an independent pressure sensor or flow sensor to collect real-time monitoring data for each segment at a fixed frequency (configurable, typically 10Hz). This real-time monitoring data includes at least the pressure value. or flow value and the pressure value or flow value The duration of the following, , Each represents a segment. i exist tPressure and flow values ​​at any given time, 1 i M. In this embodiment, a pressure sensor can be used to collect pressure values. The range of the pressure sensor can be set according to the scenario, such as 0-1.6MPa, and the actual control range (such as 0.8-1.1MPa) is adjustable.

[0039] Furthermore, this embodiment sets a dual-threshold triggering mechanism, that is, the preset triggering conditions include a first pressure threshold. The first duration threshold is used for pump anti-vibration determination. Specifically, after obtaining the segmented pressure value and the duration at that pressure value, the pressure value and the duration at that pressure value are compared with the corresponding first pressure threshold and first duration threshold, respectively. If the pressure value is less than the first pressure threshold and its duration is greater than the first duration threshold, a scheduling trigger signal is generated.

[0040] For example, in conjunction with a water usage scenario, a lower pressure limit is set for that scenario, namely a first pressure threshold. The pressure is 0.8 MPa, and the first duration threshold is 5 seconds. Therefore, if segmented... i pressure value < If the duration of this pressure value is greater than 5 seconds, then this segment is generated. i The scheduling trigger signal.

[0041] It should be understood that the aforementioned first pressure threshold Both the first duration threshold and the first duration threshold can be set according to the actual application scenario.

[0042] S300: Based on the scheduling trigger signal, extract the segmented state vector to construct the segmented decision vector.

[0043] Specifically, responding to segments i The scheduling trigger signal is used to extract the first pump in the pump group state matrix Q. i The state vector of a row ,in, Indicates the first The status of the pump, .

[0044] Based on extraction segmentation i state vector Constructing segments i Filtering vector of available pumps The construction method is as follows: (1); In formula (1): Indicates the first The vector components corresponding to the pump; if If it is 0, it means the first... The pump is in standby mode; at this time, it outputs 1, indicating that it has been selected as an available pump. If it is 1, 2, or 3, then it represents the first, second, or third. When a pump is in operation, malfunction, or idle, the output is 0, indicating that the pump is unavailable.

[0045] Subsequently, based on segmentation i Filtering vector of available pumps Combined with the random number vector R, calculate the segment. i The decision vector. This step includes: First, a K-dimensional random number vector R corresponding to each element in the vector of each segment is generated using a hardware random number generator, the Mason twitch algorithm, or a linear congruence generator. = Specifically, the elements of the random number vector The random number vector should ideally be generated using a hardware random number generator from the PLC / embedded controller. If hardware resources are unavailable, a statistically validated Mason twitch algorithm can be used to ensure the random number vector elements are valid. Independent and identically distributed in the interval (0,1) and uniformly distributed, i.e. In resource-constrained environments, a lighter-weight Linear Congruential Generator (LCG) can be used for generation.

[0046] Subsequently, the segmentation is calculated. i Filtering vector The Hadamard product with the random number vector R yields the segmentation. i Decision vector Its elements are represented as , indicating the first The decision components corresponding to the pump.

[0047] Furthermore, when counting the number N of available pumps in each segment, N is an integer and 0≤N≤K; if N=0, it means that there is no available pump in the segment, then a fault alarm is triggered and the current scheduling is terminated.

[0048] Thus, by introducing a random number vector R and combining it with the available pump selection vector... Within each scheduling cycle, a competitive set containing only N valid random numbers is dynamically constructed. Although the preset random number vector dimension is K, through the masking effect of the Hadamard product, unavailable pumps are forcibly set to zero and excluded from the competition. Therefore, for the currently available N pumps, their corresponding N independent and identically distributed random numbers in U(0,1) have an equal chance of winning in the maximum value comparison, that is, the instantaneous probability of a single pump being selected is 1 / N. This cumulative effect based on equal instantaneous probabilities mathematically guarantees that all available pumps have an equal probability of being called throughout their entire lifecycle, eliminating the "uneven wear-out" problem.

[0049] S400. Based on preset rules, select the pump index corresponding to the target component from the decision vector, and determine the pump to be scheduled based on the pump index.

[0050] In some implementations, a preset rule can be set as the maximum component rule, that is, the maximum component is selected from the decision vector as the target component to determine the pump to be scheduled. For example, assume segmentation... i Decision vector The maximum value in is the component. The value is then determined based on the component. Corresponding pump index The pump to be scheduled is determined to be the first Table pump.

[0051] In some implementations, a preset rule can be set as the minimum component rule, that is, the minimum component is selected from the decision vector as the target component to determine the pump to be scheduled. For example, assume segmentation... i Decision vector The minimum value in is a component. The value is then determined based on the component. Corresponding pump index The pump to be scheduled is determined to be the first Table pump.

[0052] Understandably, in practical applications, other selection rules can be set according to the application scenario, not limited to the maximum or minimum principles mentioned above.

[0053] Furthermore, if multiple identical components exist in the decision vector, the component with the smallest pump index is selected as the target component. For example, assume segmentation... i Decision vector There are multiple identical components, such as component If both are 0.42, then the component with the smallest pump index is selected. As the target component, the first pump can be selected as the pump to be scheduled based on its pump index 1. This ensures the uniqueness of the decision. Furthermore, the scheduling decision depends only on the pump group state matrix Q at the current moment, and is independent of historical scheduling records and historical pump group states. This fully conforms to the core definition of a Markov process: "the future state depends only on the current state." The specific connection between its mathematical mechanism and engineering implementation is as follows: Let the state matrix Q of the pump unit be the state at the current moment. , For this round of scheduling decisions, Let Q be the state of the pump group at the next time step. Then the process satisfies: (2); In formula (2): The decision function does not incorporate any historical scheduling records; This is the state transition function, which completes the transition of the system state.

[0054] Its probabilistic essence is reflected in: (3); In formula (3): This represents the historical state sequence of the pump group's state matrix Q. P The value is a probability, and its range is [0,1]. This represents the state of the pump set state matrix Q at the next time step. Based solely on the state at the current moment Decision, and the sequence of historical states This is irrelevant and fully satisfies the core definition of Markov processes having no aftereffects. This allows the system to make decisions without maintaining a sequence of historical states during computation, simply through a single operation on the current matrix, significantly reducing computational overhead.

[0055] Furthermore, after identifying each pump to be scheduled, the following steps are also included: first, start each pump to be scheduled and record the start timestamp. This is used for subsequent runtime statistics. Then, the real-time monitoring data of each segment where the scheduled pump is located is continuously monitored. If it meets the preset stop conditions, the scheduled pump with the longest runtime is shut down based on the current timestamp.

[0056] Specifically, assuming segmentation is enabled. i The Platform pump, continuous monitoring of sections i The pressure value. Preset stop conditions include a second pressure threshold. (such as ship fire-fighting equipment) =1.1 MPa, adjustable) and a second duration threshold (e.g., 60s, adjustable). Segmentation iThe pressure value and the duration at that pressure value are compared with the corresponding second pressure threshold and second duration threshold, respectively; if segmented i The pressure value is greater than or equal to the second pressure threshold. Furthermore, if the duration of operation exceeds the second duration threshold, that is, after the pressure data meets the standard and the system has been running stably for a period of time, it is determined that the preset stop conditions are met and a stop command is sent.

[0057] In some implementations, the shutdown strategy involves shutting down the pump with the longest runtime. This is the current timestamp. This strategy, combined with the aforementioned equally probable random pump start-up strategy, can jointly ensure the long-term balance of pump unit operation from both the start and stop ends.

[0058] Furthermore, after starting / stopping each scheduled pump, the state of that pump in the pump group state matrix is ​​updated accordingly. That is, immediately after a pump starts, the corresponding element in the pump group state matrix Q is updated. The status is updated from 0 (standby) to 1 (operating) and synchronized to the preset local database to ensure real-time synchronization of the pump group status. Similarly, after the pump stops, the corresponding element is updated. Update the status from 1 (running status) back to 0 (standby status), and record the duration of this run. T This data is accumulated in the pump set operating life statistics module, providing data support for subsequent maintenance. Understandably, during system operation, pump fault / repair information is uploaded in real time through the system data acquisition module or after manual confirmation, and synchronously updated to the corresponding element in the matrix pump set status matrix Q. In this matrix, the fault status is set to 2, and after repair, it is set to 0 or 1 based on the actual status, ensuring that the matrix status is completely consistent with the physical system.

[0059] Furthermore, to verify the effectiveness of this scheduling method, the following simulation comparison system was constructed: 1) Definition of comparison strategy: The commonly used axisymmetric fixed sequence pump start-up strategy (starting pumps sequentially from the middle position of the pump group number to both sides) is selected as the main comparison group to highlight the improvement of balance; it can also be compared with other intelligent scheduling algorithms.

[0060] 2) Simulation condition settings: Using the controlled variable method, under the same system parameters (such as the number of segments)... Maximum number of pumps in a single segment Actual number of pumps in each segment Under the same dynamic load sequence (such as pipeline pressure fluctuation curve) and disturbance conditions (such as pipeline leakage or pressure change), this scheduling method and the traditional FIFO (First In, First Out) scheduling method are run respectively.

[0061] 3) Selection of evaluation indicators: Select the cumulative running time of a single pump. ( k (The pump number, whose value is greater than or equal to 1 and less than or equal to the total number of pumps in the system), and the number of pump start-ups and shutdowns. , runtime standard deviation Maximum relative deviation rate As the core indicator.

[0062] Among them, standard deviation The formula used to reflect operational stability is as follows: ,in, The average of the cumulative runtime of all available pumps; relative deviation rate. The formula used to reflect the degree of wear during operation is as follows: Understandably, , The smaller the value, the better the load balance of the pump set.

[0063] 4) Data Statistics and Analysis: Set a uniform simulation cycle with a sufficient number of scheduling loops (e.g., 1000 times), and statistically analyze two types of indicators: standard deviation and relative deviation; for the relative deviation rate in the fixed sequence strategy... The high level of this phenomenon was identified as having the following cause: the inherent defect of "prioritizing the start of the intermediate pump" leads to an exponential increase in the cumulative operating time of the intermediate pump over a long period of time, ultimately causing extreme deviations.

[0064] The effectiveness of this scheduling method is further verified below with specific exemplary scenarios.

[0065] A. Exemplary Configuration (Dual Cases Across Scenarios): Case 1: Ship fire protection system (4 sections + variable pump count) a11. Determine the configuration parameters: number of segments M=4, actual number of pumps per segment: segment 1=4, segment 2=2, segment 3=3, segment 4=2; maximum number of pumps per segment K=4 (number of columns in the unified matrix). a12. Construct a 4×4 dimensional pump group state matrix Q in the initial state: the state is defined as 0 = standby, 1 = running, 2 = fault, 3 = empty (no pump configured), then: .

[0066] Case 2: Industrial Water Supply System (3-section + Variable Pump Number) a21. Determine the configuration parameters: number of segments M=3, actual number of pumps per segment: segment 1=3 units, segment 2=4 units, segment 3=2 units; maximum number of pumps per segment K=4 (number of columns in the unified matrix). a22. Construct the 3×4 dimensional pump group state matrix Q under the initial state: .

[0067] B. Example of dispatch execution (ship firefighting scenario, dispatch triggered in segment 1) b1. Triggering conditions are met: The pressure in section 1 pipeline is less than 0.7MPa (below the preset first pressure threshold of 0.7MPa) and lasts for 5 seconds, triggering the scheduling program. b2. Extract the state row vector corresponding to segment 1 in the pump group state matrix Q and filter available pumps: Extract the state row vector of segment 1: [0,1,0,2](1×4); Generate the screening vector (only standby pump = 1): [1,0,1,0](1×4); Available pump set = {pump1, pump3}, K = 2 (non-zero, continue scheduling); b3. Stochastic decision calculation: The random number vector is generated using the Mason tween rotation algorithm: [0.42, 0.59, 0.81, 0.37] (1×4); Calculate the Hadamard product (element-wise multiplication): Decision vector = [1×0.42, 0×0.59, 1×0.81, 0×0.37] = [0.42,0,0.81,0]; b4. Target Pump Selection and Execution: The maximum value of the decision vector is 0.81, corresponding to index 3 → start the 3rd pump in segment 1; Post-start monitoring: If the pipeline pressure rises to above 0.7MPa, the water supply standard is deemed met; b5. Shutdown policy execution: Segment 1: Currently operating pumps: Pump 2 (currently in operation), Pump 3 (newly started); When the pipeline pressure is detected to be greater than 1.1 MPa (higher than the preset second pressure threshold of 1.1 MPa) and lasts for 60 seconds, the scheduling program is triggered. Query cumulative runtime: Pump 2 has the longest runtime → Send a stop command to Pump 2; Update the state row vector of segment 1: [0,0,1,2] (pump 2 changes from 1 to 0, pump 3 changes from 0 to 1); b6. Synchronous update of the state matrix: ; Record the duration of this pump 2 run and synchronize it to the system database.

[0068] C. Simulation comparison and verification (ship firefighting scenario) The simulation focuses on "suppressing uneven wear and balancing the load," and is designed based on the control variable method to ensure the fairness of the comparison between the two scheduling schemes (the segmented random scheduling and the traditional FIFO scheduling in this application). The specific verification process is as follows: c1. Test Configuration and Environment Setup: Segment and Pump Set Configuration: Four independent operating segments are set up, with the actual number of pumps in each segment being 4, 2, 3, and 2 respectively, for a total of 11 pumps, which meets the requirements of variable configuration in multiple scenarios; Pressure signal control: Pressure data is generated through pure square waves, and differentiated periodic pressure curves are generated for four segments (segment 1 cycle 2700 seconds, segment 2 cycle 2400 seconds, segment 3 cycle 1800 seconds, segment 4 cycle 3600 seconds). Each cycle includes three stages: low pressure (triggered scheduling), normal, and high pressure (pump stop). The pressure signal input for segment random scheduling is completely consistent with that of FIFO scheduling, eliminating external operating condition interference. Simulation duration: Both scheduling schemes were run continuously for 1000 hours (3,600,000 seconds) with a sampling frequency of 10Hz to ensure the validity of the data statistics.

[0069] c2. Core Parameters and Indicator Settings: Dispatch trigger conditions: If the pressure is below 0.8MPa for 5 seconds (anti-vibration design), the pump will start; if the pressure rises to the target threshold (1.1MPa for segment 1 / 3, 1.15MPa for segment 2, and 1.12MPa for segment 4) and remains stable for 60 seconds, the pump will stop. Comparison metric: The relative deviation rate is selected to quantify the degree of uneven wear. and the standard deviation used to measure load dispersion As a core evaluation indicator, it directly reflects the balance of scheduling.

[0070] c3. Data Collection and Statistics: The system records the cumulative runtime and number of starts and stops of each pump under both types of scheduling in real time, and updates the pump group status matrix (standby 0 / running 1 / fault 2 / vacant 3) synchronously. Calculate the average runtime, maximum / minimum runtime, and standard deviation for each segment based on segmented statistical data. and relative deviation rate This forms a quantitative comparison dataset.

[0071] c4. Results Comparison and Analysis: Figure 3 The standard deviations of this scheduling method and the traditional FIFO scheduling method are shown. Comparison chart. As shown in the chart, the standard deviation of the segmented random scheduling method proposed in this application is only 0.1h~11.9h, the running time of each segment fluctuates very little, and the system is stable; while the standard deviation of the traditional FIFO scheduling method is 54.7h~470.8h, the segmented load distribution is chaotic, and the operation stability is poor.

[0072] Figure 4This illustrates the relative deviations between our scheduling method and the traditional FIFO scheduling method. Comparison chart. As shown in the chart, the deviation rate of the segmented random scheduling method proposed in this application is only 0.0%~4.8%, the load of each segment is highly balanced, and there is no risk of uneven wear; while the deviation rate of the traditional FIFO scheduling method is 15.0%~264.3%, and segments 1 and 3 are severely overloaded, with prominent uneven wear problems.

[0073] As can be seen from the above, this scheduling method, by introducing the Hadamard product and combining it with an equal-probability random decision-making mechanism, achieves equal probability of pump dispatch for each pump, and its relative deviation rate... Only 0.0%~4.8% of the standard deviation With a time of only 0.1h to 11.9h, it is far superior to FIFO scheduling (relative deviation rate 15.0% to 264.3%, standard deviation 54.7h to 470.8h), fundamentally solving the problems of uneven wear and load in traditional FIFO scheduling.

[0074] Unlike traditional fixed-sequence strategies where deviations grow linearly or exponentially over time, this scheduling method employs an equal-probability random strategy, causing the difference in cumulative runtime of each pump to exhibit a random walk, with an expected value of zero and a standard deviation of zero. As the square root of time increases, the load unevenness can be controlled to an acceptable low level during long-term operation.

[0075] In summary, this scheduling method, by introducing a state matrix and a missing state isolation mechanism, achieves standardized and unified management of any number of segments and any number of pumps in a single segment, breaking the limitations of scenario and physical configuration. At the same time, it formulates scheduling strategies based on real-time monitoring data of segments, ensuring that all available pumps have an equal probability of being called in the long term, fundamentally eliminating the problem of uneven wear, extending the overall service life of the equipment, and solving the core problems of poor universality, easy uneven wear, and difficult engineering deployment of traditional pump group scheduling.

[0076] Secondly, this embodiment provides a distributed pump set anti-wear scheduling system, which applies the distributed pump set anti-wear scheduling method described above. Please refer to [link / reference]. Figure 5 The scheduling system includes: The matrix construction module 10 is used to determine the number of segments and the maximum number of pumps in a single segment, construct the pump group state matrix, and initialize the pump group state matrix. The signal generation module 20 is used to acquire real-time monitoring data of each segment in the pump group status matrix. When the real-time monitoring data of a segment meets the preset trigger conditions, it generates a scheduling trigger signal for the segment. The vector calculation module 30 is used to extract the segmented state vectors based on the scheduling trigger signal, in order to construct the segmented decision vectors; The scheduling selection module 40 is used to select the pump index corresponding to the target component from the decision vector based on preset rules, and determine the pump to be scheduled based on the pump index.

[0077] Further, please refer to Figure 6 The dispatching system also includes a water intake module and a pipeline module. The water intake module includes a submarine gate, seawater filter, etc., to provide a stable water source for the system. The pipeline module includes a main pipe and zone valve groups, water terminals (cooling, fire protection, decontamination, domestic use), etc., to realize fluid transportation.

[0078] Furthermore, this system can be directly adapted to commonly used industrial hardware such as PLC controllers and embedded modules without additional upgrades or modifications, resulting in a low barrier to engineering deployment. Based on a generalized architecture design, it possesses strong cross-domain migration capabilities and can be seamlessly extended to balanced scheduling scenarios involving multiple parallel devices such as multi-generator clusters, industrial AGV robot groups, and charging pile clusters. It provides a standardized, quantifiable, and easily implementable general technical solution to address the common problems of "uneven wear and tear, limited configuration, and complex deployment" in multi-device clusters.

[0079] Thirdly, this embodiment also provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, it implements the distributed pump group anti-wear scheduling method as described above.

[0080] Fourthly, this embodiment also provides a computer-readable storage medium storing a computer program thereon, which is loaded by a processor to execute the steps in the distributed pump group anti-wear scheduling method described above.

[0081] In the embodiments of this application, the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM), etc.

[0082] It should be noted that, in the data processing stage, the technical solution of this application has strictly limited the scope of data collection to the minimum necessary to achieve the technical objectives, preventing the acquisition of irrelevant information. For any user information to be collected, the data subject will be clearly informed and their consent obtained. Furthermore, technologies such as encrypted storage and access control are employed to strengthen data security and ensure the security and compliance of the entire data processing process. The technical model and decision-making mechanism are based on objective technical parameters and do not introduce unnecessary parameters such as gender or age that may lead to discrimination, resolutely eliminating algorithmic discrimination and upholding public order and good morals. In addition, the specification fully describes the technical implementation methods, application scenarios, and compliance protection details. The claims are consistent with the content of the specification, key compliance designs are clear and verifiable, and the overall technical design is guided by the protection of public interests and adherence to social ethics, without any circumstances that harm public interests or violate public order and good morals.

[0083] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0084] The above provides a detailed description of a distributed pump group anti-wear scheduling method, system, electronic device, and storage medium provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for scheduling distributed pump sets to prevent uneven wear, characterized in that, Includes the following steps: Determine the number of segments and the maximum number of pumps per segment, and construct the pump group state matrix; The real-time monitoring data of each segment in the pump group status matrix is ​​obtained, and when the real-time monitoring data of the segment meets the preset triggering conditions, the scheduling trigger signal of the segment is generated. Based on the scheduling trigger signal, the state vector of the segment is extracted to construct the decision vector of the segment; Based on preset rules, the pump index corresponding to the target component is selected from the decision vector, and the pump to be scheduled is determined based on the pump index.

2. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, The step of extracting the state vector of the segment to construct the decision vector of the segment includes: Based on the state vector of the segment, a selection vector for the available pumps of the segment is constructed; The decision vector is calculated based on the filtering vector and combined with the random number vector.

3. The distributed pump set anti-wear scheduling method according to claim 2, characterized in that, The random number vector is generated based on a hardware random number generator, the Mason tween algorithm, or a linear congruence generator.

4. The distributed pump set anti-wear scheduling method according to claim 2, characterized in that, The process of constructing the selection vector for available pumps in the segment also includes: If the number of available pumps in the segment is zero, a fault alarm is triggered and the current scheduling is terminated.

5. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, After constructing the pump set state matrix, the method further includes: The pump group state matrix is ​​initialized based on the encoded values ​​corresponding to the initial states of all pumps. The initial state of the pump includes standby state, running state, and fault state.

6. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, After constructing the pump set state matrix, the method further includes: If the actual number of pumps in a segment is less than the maximum number of pumps in a single segment, the element in the pump group state matrix corresponding to the segment without a pump is initialized to a missing code value.

7. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, The real-time monitoring data for each segment includes the pressure value and the duration of that pressure value; The preset triggering conditions include a first pressure threshold and a first duration threshold; The generation of the segmented scheduling trigger signal includes: The pressure value of the segment and the duration of the pressure value are compared with the corresponding first pressure threshold and first duration threshold, respectively. If the pressure value of the segment is less than the first pressure threshold and its duration is greater than the first duration threshold, a scheduling trigger signal is generated.

8. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, The preset rule is the maximum component rule; The step of selecting the pump index corresponding to the target component from the decision vector based on preset rules includes: The largest component is selected from the decision vector as the target component, and the pump to be scheduled is determined based on its corresponding pump index.

9. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, The preset rule is the minimum component rule; The step of selecting the pump index corresponding to the target component from the decision vector based on preset rules includes: The minimum component is selected from the decision vector as the target component, and the pump to be scheduled is determined based on its corresponding pump index.

10. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, The step of selecting the pump index corresponding to the target component from the decision vector based on preset rules further includes: If there are multiple identical components in the decision vector, the component with the smallest pump index is selected as the target component.

11. The distributed pump set anti-wear scheduling method according to claim 1, characterized in that, After determining each pump to be scheduled, the method further includes: Start each of the scheduled pumps and record the start timestamp; The real-time monitoring data of each segment containing the pump to be scheduled is continuously monitored. If the preset stop conditions are met, the pump with the longest running time is shut down based on the current timestamp.

12. The distributed pump set anti-wear scheduling method according to claim 11, characterized in that, The segmented real-time monitoring data includes the pressure value and the duration of that pressure value; The preset stop conditions include a second pressure threshold and a second duration threshold; Determining whether the preset stop condition is met includes: The pressure value of the segment and the duration of the pressure value are compared with the corresponding second pressure threshold and second duration threshold, respectively. If the pressure value of the segment is greater than or equal to the second pressure threshold and its duration is greater than the second duration threshold, then the preset stop condition is satisfied.

13. The distributed pump set anti-wear scheduling method according to claim 11, characterized in that, After starting / stopping each of the scheduled pumps, the method further includes: The status of the pump to be scheduled in the pump group status matrix is ​​updated accordingly.

14. A distributed pump set anti-wear scheduling system, characterized in that, The distributed pump set anti-wear scheduling method as described in any one of claims 1-13, wherein the scheduling system comprises: The matrix construction module is used to determine the number of segments and the maximum number of pumps per segment, construct the pump group state matrix, and initialize the pump group state matrix. The signal generation module is used to acquire real-time monitoring data of each segment in the pump group status matrix, and generate a scheduling trigger signal for the segment when the real-time monitoring data of the segment meets the preset trigger conditions. The vector calculation module is used to extract the state vector of the segment based on the scheduling trigger signal, so as to construct the decision vector of the segment; The scheduling selection module is used to select the pump index corresponding to the target component from the decision vector based on preset rules, and determine the pump to be scheduled based on the pump index.

15. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the distributed pump group anti-wear scheduling method as described in any one of claims 1-13.

16. A computer-readable storage medium, characterized in that, It stores a computer program, which is loaded by a processor to execute the steps in the distributed pump set anti-wear scheduling method as described in any one of claims 1-13.

Images