Multi-objective optimization of heat station primary network and secondary pump operation control method

Abstract

The present application relates to the technical field of hot water hydraulic regulation, in particular to a multi-objective optimization heat station primary network secondary pump operation regulation method. In the present application, by constructing a primary network hydraulic topology and analyzing pressure consistency deviation, a hydraulic feature center pipeline representing the overall potential energy level of the region is accurately positioned, control lag caused by local singular points is avoided, a load constraint boundary generated by mapping three-dimensional operation data of pump frequency, flow and power is used, a frequency domain flow reachable interval is delimited in combination with a pipeline network resistance characteristic curve, pump operation is strictly limited within a safe and efficient envelope surface, based on multi-objective optimization logic balancing pressure tracking accuracy and power safety margin, the best frequency instruction is dynamically solved, accurate matching of energy supply and pipeline network demand is realized, and the hydraulic stability and operation energy efficiency ratio of the heating system are significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of heat supply hydraulic regulation, in particular to a multi-objective optimization heat station primary network secondary pump operation regulation method. BACKGROUND

[0002] The technical field of heat supply hydraulic regulation relates to the process control of heat supply medium in the pipe network in the central heating system, the core matters of which include the hydraulic balance regulation between the heat source side and the user side, the coordinated distribution of pipe network supply and return water pressure and flow, and the regulation strategy of circulating water pump operation mode, and the systematic management around the heat exchange station, the primary network and the secondary network hydraulic coupling relationship, so that the heat supply system can maintain the ability to stably deliver heat under different loads and different working conditions.

[0003] Among them, the heat station primary network secondary pump operation regulation method usually sets the start and stop conditions of the secondary pump according to the primary network supply and return water temperature and pressure, realizes the matching of the delivery flow by fixedly selecting the number of pump groups, and runs the secondary pump in a constant speed or multi-speed mode; when regulating the heat exchange link, the set opening throttle valve or the preset valve opening is usually used to control the secondary network side flow; when maintaining the pipe network circulation, the fixed pressure difference control point is usually set and maintained by the secondary pump to maintain the secondary network circulation; under the condition of load change, the number of running secondary pumps or the speed gear is usually adjusted according to the change range of the primary network pressure or temperature to complete the regulation.

[0004] The existing technology simply relies on fixed pressure difference control point to regulate pipe network circulation, ignores the overall hydraulic coupling effect of supply and return water pipe network, causes local pressure imbalance and uneven heat distribution, uses extensive pump group number increase and decrease or multi-speed regulation mode, cannot accurately match the continuously changing pipe network hydraulic resistance characteristics, causes serious invalid energy consumption under part load working condition, lacks dynamic correlation mechanism between pump motor power and actual hydraulic demand, increases the risk of motor overload or deviating from the high efficiency operation interval, relies on throttle valve to control flow, introduces additional hydraulic impedance loss, and limits the improvement of overall energy efficiency ratio of heat supply delivery process. SUMMARY

[0005] In order to solve the technical problems existing in the prior art, the multi-objective optimization heat station primary network secondary pump operation regulation method provided by the embodiments of the present application comprises the following steps:

[0006] In order to achieve the above purpose, the technical scheme adopted by the present application is as follows: the multi-objective optimization heat station primary network secondary pump operation regulation method comprises the following steps:

[0007] S1: constructing a primary network hydraulic connection topology according to a physical connection sequence of a supply and return water pipeline of a heat station primary network and a valve opening and closing state, and dividing a hydraulic control partition, and counting a partition pipeline pressure data set of each hydraulic control partition;

[0008] S2: filtering a hydraulic feature center pipeline according to the partition pipeline pressure data set and extracting a center pipeline real-time pressure;

[0009] S3: synchronously acquiring a secondary pump running frequency of a secondary pump, a primary network pipeline water flow and a secondary pump motor real-time power at a current moment, constructing a three-dimensional running condition data set, and dividing a three-dimensional power load constraint boundary through a preset safety judgment logic;

[0010] S4: constructing a primary network pipeline resistance characteristic curve based on a relationship between the current primary network pipeline water flow and the center pipeline real-time pressure, scanning an intersection position of the curve and the three-dimensional power load constraint boundary, and constructing a frequency domain flow reachable interval boundary line;

[0011] S5: taking a deviation minimization of the center pipeline real-time pressure and a primary network heating target pressure as a first optimization target, and taking a current running point of the secondary pump and the frequency domain flow reachable interval boundary line as a second optimization target, and generating a regulation and control frequency instruction.

[0012] As a further scheme of the application, the partition pipeline pressure data set comprises a hydraulic control partition code divided according to a physical connection, a supply and return water pipeline position index mapped to each partition, and a real-time pressure monitoring value corresponding to each index, the center pipeline real-time pressure comprises a hydraulic feature center pipeline identification filtered according to a consistency offset and a current real-time pressure measurement reading of the pipeline, the three-dimensional power load constraint boundary specifically comprises a discrete frequency sampling sequence under a rated power condition and a maximum primary network pipeline water flow extreme value corresponding to each sampling frequency point, the frequency domain flow reachable interval boundary line specifically comprises a spatial intersection point coordinate of the three-dimensional power load constraint boundary and the primary network pipeline resistance characteristic curve and a flow physical limit contour formed based on the intersection point connection, and the regulation and control frequency instruction comprises a frequency converter target running frequency setting instruction and a speed regulation execution instruction for driving the secondary pump action.

[0013] As a further scheme of the application, step S1 specifically comprises:

[0014] S101: acquiring a physical connection sequence of a supply and return water pipeline of a heat station primary network and a valve opening and closing state, establishing a pipeline node connection relationship according to the physical connection sequence, combining the valve opening and closing state to determine a conduction condition between pipelines, and generating a primary network hydraulic connection topology;

[0015] S102: retrieve the current state of all valves in the primary network hydraulic connection topology, identify the valves in the closed state and mark as physical isolation boundary, according to the physical isolation boundary, the overall connected topology is disassembled into independent subnets that are not connected to each other, determine the range of pipes covered by each independent subnet, and generate independent hydraulic control partition;

[0016] S103: using the pressure sensor arranged in each pipe position to collect the real-time pressure of the primary network supply and return pipe at the same time, obtaining the physical position information of the pressure sensor, matching the collected pressure value to the corresponding independent hydraulic control partition range according to the physical position information, and generating a partition pipe pressure data set.

[0017] As a further scheme of the present application, step S2 is specifically:

[0018] S201: traversing all real-time pressure data of primary network supply and return water pipes in the partition pipe pressure data set, calculating the regional average pressure value of the overall pressure level in the current hydraulic control partition;

[0019] S202: extracting the real-time pressure of each pipe position in the partition pipe pressure data set one by one, calculating the comparison data representing the difference between the two values according to the real-time pressure of each pipe position and the regional average pressure value, quantifying the deviation of each single-point pressure data relative to the regional average level, and generating a pressure consistency offset;

[0020] S203: arranging all the pressure consistency offsets in order from small to large, locking the minimum offset value at the first position, tracing back to the primary network pipe physical position corresponding to the minimum offset data, marking it as the hydraulic feature center pipe representing the overall hydraulic state of the partition, and extracting the real-time pressure of the center pipe.

[0021] As a further scheme of the present application, step S3 is specifically:

[0022] S301: synchronously collecting the secondary pump rated frequency range, the secondary pump operating frequency, the primary network pipe network water flow and the secondary pump motor real-time power at the current time, establishing a frequency-flow-power three-dimensional coordinate space with three dimensions of frequency, flow and power, mapping the three collected data into coordinate points in the space, and generating a three-dimensional operating condition data set;

[0023] S302: comparing the real-time power of the secondary pump motor in the three-dimensional operating condition data set with the preset rated power reference, eliminating abnormal data exceeding the power limit, traversing and retrieving the flow record at each frequency point within the secondary pump rated frequency range, extracting the maximum primary network pipe network water flow value under the condition of meeting the power safety limit, and generating a non-overload maximum flow node;

[0024] S303: remap the non-overloaded maximum flow node back into the frequency-flow-power three-dimensional coordinate space, connect the maximum flow and corresponding power data at each frequency point to form a continuous spatial surface or geometric envelope, establish the power safety range of the secondary pump under multiple operating frequency and flow combinations, and generate a three-dimensional power load constraint boundary.

[0025] As a further scheme of the present application, the process of establishing a frequency-flow-power three-dimensional coordinate space with three dimensions of frequency, flow and power is specifically:

[0026] A spatial rectangular coordinate system is constructed with the operating frequency of the secondary pump, the water flow of the primary network pipe and the real-time power of the secondary pump motor as independent coordinate axes.

[0027] The three data obtained synchronously at the same time are associated and bound, and are defined as a spatial feature vector describing the state at a single operating time.

[0028] According to the numerical values of the components of the spatial feature vector, a unique geometric coordinate point is located in the spatial rectangular coordinate system, and all geometric coordinate points under continuous time series are collected and stored.

[0029] As a further scheme of the present application, step S4 is specifically:

[0030] S401: According to the comparison relationship between the water flow of the primary network pipe and the real-time pressure of the center pipe in the three-dimensional operating condition data set, the hydraulic impedance characteristic of the current pipe network is quantified, fitted as a trajectory line reflecting the physical state and flow characteristic of the current pipe network, and a primary network pipe resistance characteristic curve is generated.

[0031] S402: Set a frequency domain scanning interval covering the rated operating range of the secondary pump, and discretize the frequency domain scanning interval into multiple discrete scanning frequency points using a preset step size. In the frequency-flow-power three-dimensional coordinate space, the contact position of the primary network pipe resistance characteristic curve and the three-dimensional power load constraint boundary under each frequency point is determined one by one, the flow value of the spatial intersection point of the contact position is calculated, and the physical limit flow under each frequency is locked.

[0032] S403: Map the physical limit flow corresponding to all discrete frequency points to the frequency domain analysis plane, logically connect each limit flow point in order of increasing frequency, and demarcate the operating range of the secondary pump under the conditions of power safety and pipe network resistance, and generate a frequency domain flow reachable interval boundary line.

[0033] As a further scheme of the present application, the process of determining the contact position of the primary network pipe resistance characteristic curve and the three-dimensional power load constraint boundary under each frequency point in the frequency-flow-power three-dimensional coordinate space is specifically:

[0034] For each discrete scanning frequency point, in the frequency-flow-power three-dimensional coordinate space, find the intersection state point that meets the function constraint of the first-level network management network resistance characteristic curve and the spatial geometric definition of the three-dimensional power load constraint boundary at the same time;

[0035] Analyze the projection value of the intersection state point on the flow coordinate axis to establish the physical limit flow under the corresponding frequency.

[0036] As a further scheme of the present application, step S5 is specifically:

[0037] S501: Obtain the preset first-level network heating target pressure, combine the real-time pressure of the central pipeline, calculate the numerical difference between the two to establish the optimization direction of the pressure tracking accuracy, and simultaneously call the current running state coordinates of the secondary pump from the three-dimensional running condition data set, calculate the Euclidean distance between the running state coordinates in the frequency-flow-power three-dimensional space projection plane and the frequency domain flow reachable interval boundary line, establish the optimization direction of the power safety margin, and generate a multi-objective optimization function;

[0038] S502: Set the frequency domain flow reachable interval boundary line as a hard constraint condition, input the multi-objective optimization function into the NSGA-II multi-objective optimization algorithm, perform multiple rounds of iterative optimization calculation, evaluate the function performance under multiple frequency settings in the global search process, eliminate solutions that do not meet the boundary constraint, locate the best working condition configuration, and extract the optimal frequency value;

[0039] S503: Map the optimal frequency value to a standardized signal suitable for the control logic of the frequency converter, analyze the frequency adjustment amplitude and direction information of the first-level network secondary pump of the heating station, and generate a frequency control instruction.

[0040] As a further scheme of the present application, the process of generating a multi-objective optimization function is specifically:

[0041] A first evaluation dimension is constructed to minimize the absolute value of the difference between the real-time pressure of the central pipeline and the first-level network heating target pressure, which measures the accuracy of pressure tracking;

[0042] A second evaluation dimension is constructed to maximize the spatial distance from the current running state coordinates of the secondary pump to the frequency domain flow reachable interval boundary line, which measures the safety margin of the running condition;

[0043] The first evaluation dimension and the second evaluation dimension are logically combined to form a multi-objective optimization function that quantitatively evaluates the pressure deviation and safety distance at the same time.

[0044] Compared with the prior art, the present application has the following advantages and positive effects:

[0045] In the application, by constructing a primary network hydraulic topology and analyzing pressure consistency deviation, the hydraulic feature center pipeline representing the overall potential energy level of the region is accurately positioned, the control lag caused by local singular points is avoided, the load constraint boundary generated by mapping the three-dimensional operation data of pump frequency, flow and power is used, the frequency domain flow reachable interval is delimited combined with the pipe network resistance characteristic curve, the pump operation is strictly limited within the safe and efficient envelope surface, the optimal frequency instruction is dynamically solved based on the multi-objective optimization logic of balancing pressure tracking accuracy and power safety margin, the accurate matching of energy supply and pipe network demand is realized, and the hydraulic stability and operation energy efficiency ratio of the heating system are significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0047] Figure 1 The step flowchart of the present application is shown in the figure.

[0048] Figure 2 The S1 refinement schematic diagram of the present application is shown in the figure.

[0049] Figure 3 The S2 refinement schematic diagram of the present application is shown in the figure.

[0050] Figure 4 The S3 refinement schematic diagram of the present application is shown in the figure.

[0051] Figure 5 The S4 refinement schematic diagram of the present application is shown in the figure.

[0052] Figure 6 The S5 refinement schematic diagram of the present application is shown in the figure. DETAILED DESCRIPTION

[0053] The technical solutions in the present application will be described below in combination with the drawings.

[0054] In the embodiments of the present application, the words such as "example", "for example" are used to represent as an example, illustration or description. Any embodiment or design scheme described as "example" in the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present application, the meaning expressed by "and / or" can be both, or can be one of the two.

[0055] In order to make the technical problems, technical solutions and advantages of the present application clearer, the following will be described in detail in combination with the drawings and specific embodiments.

[0056] Referring to Figure 1 The embodiment of the present application provides a multi-objective optimization heat station primary network secondary pump operation regulation method, comprising the following steps:

[0057] S1: According to the physical connection sequence of the heat station primary network supply and return water pipeline and the valve opening and closing state, a primary network hydraulic communication topology structure is constructed, and a hydraulic control partition is divided, and a partition pipeline pressure data set of each hydraulic control partition is counted;

[0058] S2: According to the partition pipeline pressure data set, a hydraulic characteristic central pipeline is screened and the real-time pressure of the central pipeline is extracted;

[0059] S3: The secondary pump operation frequency of the secondary pump, the primary network pipeline water flow and the real-time power of the secondary pump motor at the current time are synchronously acquired, a three-dimensional operation condition data set is constructed, and a three-dimensional power load constraint boundary is divided through a preset safety judgment logic;

[0060] S4: A primary network pipeline resistance characteristic curve is constructed based on the relationship between the current primary network pipeline water flow and the real-time pressure of the central pipeline, the intersection position of the curve and the three-dimensional power load constraint boundary is scanned, and a frequency domain flow reachable interval boundary line is constructed;

[0061] S5: Taking the minimization of the deviation between the real-time pressure of the central pipeline and the primary network heating target pressure as the first optimization target, and taking the current operation point of the secondary pump and the frequency domain flow reachable interval boundary line as the second optimization target, a regulation frequency instruction is generated.

[0062] The partition pipeline pressure data set includes the hydraulic control partition code divided according to the physical connection, the supply and return water pipeline position index mapped to each partition, and the real-time pressure monitoring value corresponding to each index. The real-time pressure of the central pipeline includes the hydraulic characteristic central pipeline identifier screened according to the consistency offset, and the current real-time pressure measurement reading of the pipeline. The three-dimensional power load constraint boundary is specifically a discrete frequency sampling sequence that meets the rated power condition, and the maximum primary network pipeline water flow extreme value corresponding to each sampling frequency point. The frequency domain flow reachable interval boundary line is specifically the spatial intersection point coordinates of the three-dimensional power load constraint boundary and the primary network pipeline resistance characteristic curve, and the flow physical limit contour formed based on the intersection point connection. The regulation frequency instruction includes a frequency converter target operation frequency setting instruction and a speed regulation execution instruction for driving the secondary pump to act.

[0063] Referring to Figure 2 Step S1 is specifically:

[0064] S101: Obtain the physical connection sequence of the supply and return water pipes of the heat station primary network and the opening and closing states of the valves, establish the pipe node connection relationship according to the physical connection sequence, determine the conduction between the pipes in combination with the opening and closing states of the valves, and generate a primary network hydraulic connectivity topology structure;

[0065] The pipe network basic data in the geographic information library is called, the pre-recorded pipe unique identification code, pipe starting coordinates, pipe ending coordinates and physical interface form between pipes are read, discrete pipe data is spliced into continuous physical links according to the coordinate coincidence degree and interface matching, the real-time feedback signals of the electric regulating valves and shut-off valves installed at each key node of the pipe network are read through the data acquisition and monitoring control device, the feedback signals include Boolean state bits or analog opening degree values. For the analog opening degree value, a preset conduction determination threshold is called for comparison. The setting basis of the conduction determination threshold is: the flow leakage data of the same type of valve at different opening degrees is counted, the average value of the minimum opening degree at which the flow begins to produce substantial conduction is calculated, and the average value is set as the conduction determination threshold (for example, if the average flow starting opening degree is 3%, the threshold is set to 3%). If the valve real-time feedback opening degree value is greater than 3% or the feedback state bit signal is on, the valve position is determined as a conduction node. In establishing the connection relationship, an adjacency matrix storage method is used, each pipe is regarded as a graph vertex, and the physical connection between the pipes is regarded as an edge. In combination with the valve state, if the valve connecting two pipes is in a conduction state, the corresponding position in the adjacency matrix is marked as a connected identifier; if the valve is in a closed state, it is marked as a disconnected identifier. All pipe nodes and valve nodes are traversed, and the primary network hydraulic connectivity topology structure reflecting the actual flow direction path of the water flow at the current time is constructed according to the above logic.

[0066] S102: Retrieve the current state of all valves in the primary network hydraulic connectivity topology structure, identify the valves in the closed state and mark them as physical isolation boundaries, decompose the overall connectivity topology structure into independent sub-networks that are not connected to each other according to the physical isolation boundaries, determine the pipe range covered by each independent sub-network, and generate independent hydraulic control partitions;

[0067] The current state of all valves in the primary network hydraulic connection topology is searched, and valves with an opening degree feedback value less than or equal to the on-off threshold value or with a state bit showing off are identified as physical isolation boundaries that block water flow transmission. A connection component search operation is performed, taking any unvisited pipe node as a starting point, traversing along the paths marked as connected in the adjacency matrix until a physical isolation boundary or no subsequent connection node is encountered, collecting all pipe nodes that can be visited in this traversal into a set, and confirming that the set is an independent and unconnected subnetwork. The above search process is repeated until all pipe nodes in the topology structure are assigned to a specific subnetwork set. For each determined independent subnetwork, the geographical location coordinate range and pipe length information of all pipes in the set are read, and the actual coverage area of the subnetwork in the geographical space is circled, and the area is defined as an independent hydraulic control partition. If there are two closed connection valves in the pipe network of a heat station, the entire pipe network is divided into three independent operation areas according to the two breakpoints, and three independent hydraulic control partition data packets containing specific pipe lists are output.

[0068] S103: Collecting real-time pressures of the primary network supply and return pipes at the same time by using pressure sensors arranged at each pipe location, obtaining physical location information of the pressure sensors, matching the collected pressure values to the corresponding independent hydraulic control partition range according to the physical location information, and generating a partition pipe pressure data set;

[0069] Collecting real-time pressures of the primary network supply and return pipes at the same time by using pressure sensors arranged at each pipe location. The current signals of each pressure transmitter are read by polling through an industrial field bus protocol (usually 4 to 20 milliamps). The sensor range parameters stored in the device parameter library are called, such as a minimum range of 0 megapascals and a maximum range of 1.6 megapascals. The real-time pressure value is calculated using a linear conversion formula. For example, if the collected real-time current is 12 milliamps, the real-time pressure calculation process is: , and the calculation result is 0.8 megapascals. At the same time, the installation location coordinates or the bound pipe identification code of each pressure sensor are extracted from the device account database. The data mapping operation is performed to associate the calculated pressure value with the corresponding pipe identification code, and then the pipe identification code is compared with the generated pipe list in the independent hydraulic control partition. If the pipe identification code corresponding to a certain pressure sensor is included in the pipe list of a specific partition, the real-time pressure value of the sensor is included in the data set of the partition. All sensor data are traversed to complete the classification processing, and a partition pipe pressure data set for each independent hydraulic control partition is generated.

[0070] Please refer to Figure 3 , step S2 is specifically:

[0071] S201: Traverse all real-time pressure data of primary water supply pipeline in the partition pipeline pressure data set, calculate the regional average pressure value of the overall pressure level in the current hydraulic control partition;

[0072] First, count the total number of effective pressure data in the current independent hydraulic control partition, then add up all the real-time pressure values in the set. Divide the total pressure sum obtained by the total number of data to obtain the regional average pressure value of the overall pressure level in the current hydraulic control partition. For example, if there are 5 effective pressure monitoring points in the partition, their pressure values are 0.50 MPa, 0.52 MPa, 0.48 MPa, 0.51 MPa and 0.49 MPa. Calculate the pressure sum: . Calculate the regional average pressure value: The calculation result is 0.50 MPa. This value eliminates the local fluctuation interference that may exist in single-point measurement, and reflects the macroscopic average potential energy level of the closed hydraulic region under the current working condition.

[0073] S202: Extract the real-time pressure of each pipeline position in the partition pipeline pressure data set one by one, calculate the comparison data representing the difference between the two values according to the real-time pressure of each pipeline position and the regional average pressure value, quantify the deviation of each single-point pressure data relative to the regional average level, and generate the pressure consistency offset;

[0074] Extract the real-time pressure of each pipeline position in the partition pipeline pressure data set one by one, take it as the minuend, take the calculated regional average pressure value as the subtrahend, and perform difference operation. Take the absolute value of the difference to eliminate the directional influence and only retain the amplitude of the numerical difference, and generate the pressure consistency offset. The offset value exists in multiple possible intervals: if the offset belongs to the interval MPa, it is determined as a high consistency interval; if the offset belongs to the interval MPa, it is determined as a medium consistency interval; if the offset belongs to the interval MPa, it is determined as a low consistency interval. The setting basis of the interval limit value is: calculate the standard deviation of the pressure of each monitoring point under the stable running state of the pipeline network in history, take one standard deviation as the upper limit of high consistency, and take three standard deviations as the upper limit of medium consistency (for example, if the standard deviation is calculated to be 0.01 MPa, set the values around 0.01 and 0.03, and here the example takes 0.05 as the upper limit of the medium interval). For example, for a monitoring point with a real-time pressure of 0.52 MPa and a regional average pressure of 0.50 MPa, calculate the absolute value of the difference: The calculation result is 0.02 MPa. The value falls in In the interval, it is determined to belong to the medium consistency interval. For the measuring point with a real-time pressure of 0.50 MPa, the offset is 0, which belongs to the high consistency interval. The above calculation is performed on each measuring point in the set to generate a set of pressure consistency offset data sequences corresponding to the measuring points.

[0075] S203: arrange all pressure consistency offsets in order from small to large, lock the minimum offset value at the first position, trace back the physical position of the primary network pipeline corresponding to the minimum offset data, mark the center pipeline representing the overall hydraulic state of the partition as a hydraulic feature, and extract the real-time pressure of the center pipeline;

[0076] All pressure consistency offsets are arranged in order from small to large. The value at the first position after sorting, that is, the minimum offset value, is locked. If there are multiple identical minimum offsets, trace back the topological distance of the pipeline nodes from the heat source inlet corresponding to these offsets, and select the node with the shortest topological distance. Trace back the original pressure sensor identification code corresponding to the minimum offset data, and further search for the physical position of the primary network pipeline where the sensor is installed in the pipeline network geographic information database. Mark this physical position as the hydraulic feature center pipeline. Then, directly extract the real-time pressure reading of the sensor at the current sampling period as the center pipeline real-time pressure representing the overall hydraulic state of the partition. For example, in the above data, the measuring point with a pressure of 0.50 MPa has an offset of 0, which is the minimum value, and the pipeline where the sensor is installed is set as the center pipeline, and 0.50 MPa is extracted as the core feedback variable for subsequent control.

[0077] Please refer to Figure 4 , step S3 is specifically:

[0078] S301: synchronously collect the secondary pump rated frequency range, the secondary pump operating frequency, the primary network pipeline water flow, and the secondary pump motor real-time power at the current time, establish a frequency-flow-power three-dimensional coordinate space with three dimensions of frequency, flow, and power, map the three collected data to coordinate points in the space, and generate a three-dimensional operating condition data set;

[0079] The process of establishing a frequency-flow-power three-dimensional coordinate space with three dimensions of frequency, flow, and power is specifically:

[0080] Construct a space rectangular coordinate system with the secondary pump operating frequency, the primary network pipeline water flow, and the secondary pump motor real-time power as independent coordinate axes;

[0081] Correlate and bind the three data synchronously acquired at the same time, and define it as a space feature vector describing the state at a single operating time;

[0082] According to the component values of the spatial feature vector, a unique geometric coordinate point is located in the spatial rectangular coordinate system, and all geometric coordinate points under continuous time series are collected and stored;

[0083] The running state data of the secondary pump at the current time is synchronously collected. The current running frequency of the secondary pump is read through the communication interface of the frequency converter; the instantaneous water flow through the primary network is read through the electromagnetic flowmeter; the real-time active power of the secondary pump motor is read through the power analyzer. At the same time, the rated frequency range (for example, 0 to 50 Hz) of the secondary pump stored in the device parameter library is called. A three-dimensional rectangular coordinate system is constructed, and frequency, flow, and power are used as independent coordinate axes respectively. The frequency value, flow value, and power value obtained at the same sampling time are combined into an array consisting of three values, which is defined as a spatial feature vector. According to the three component values of the vector, a unique spatial geometric point is determined in the three-dimensional coordinate system. According to the preset sampling time interval (for example, 10 seconds), these geometric points are continuously stored into the historical database to form a three-dimensional running condition data set containing a large number of historical running state points. The setting basis of the preset sampling time interval is: the response delay time of the flow after the frequency regulation of the secondary pump is counted, and half of the response delay time is taken as the sampling interval (for example, the response delay is 20 seconds, and the sampling interval is set to 10 seconds) to ensure that the dynamic change process is captured.

[0084] S302: Compare the real-time power of the secondary pump motor in the three-dimensional running condition data set with the preset rated power reference, eliminate abnormal data exceeding the power limit, traverse and retrieve the flow record at each frequency point within the rated frequency range of the secondary pump, extract the maximum primary network water flow value reached under the power safety limit, and generate a non-overload maximum flow node;

[0085] The rated power value (for example, 55 kW) on the nameplate of the secondary pump motor is read, and a safety power coefficient is set. The setting basis of the safety power coefficient is: the power critical point of insulation aging caused by overheating in the historical operation of the motor is counted, the ratio of the critical point to the rated power is calculated, and 95% of the ratio is taken as the safety power coefficient (for example, the critical point is counted as 58 kW, the ratio is calculated as , and 0.95 is taken as the coefficient to leave a margin). The upper limit of power safety is calculated: , the calculation result is 52.25 kilowatts. Traverse each data point in the three-dimensional operating condition data set, compare the recorded real-time power with the upper limit of the power safety. If the power value of a certain data point exceeds 52.25 kilowatts, it is determined that the point is an overload operating condition, and it is excluded from the data set. For the remaining compliant data, in the rated frequency range of the secondary pump, a plurality of frequency intervals are divided at a preset frequency step (for example, 0.5 hertz). In each frequency interval, all historical flow records are retrieved, and the corresponding maximum flow value at the frequency is found by numerical comparison. Each frequency point and its corresponding maximum flow value are combined to generate a series of non-overload maximum flow nodes.

[0086] S303: remap the non-overload maximum flow nodes back to the frequency-flow-power three-dimensional coordinate space, and connect the maximum flow at each frequency point with the corresponding power data to form a continuous spatial surface or geometric envelope, to determine the power safety range of the secondary pump under multiple operating frequencies and flow combinations, and generate a three-dimensional power load constraint boundary;

[0087] The screened non-overload maximum flow nodes are remapped back to the frequency-flow-power three-dimensional coordinate space. These nodes appear as a series of scattered points in space, and a least squares surface fitting algorithm is used to construct a continuous spatial surface based on these discrete points. This surface describes the critical state of the secondary pump under different operating frequencies and flow combinations when the power reaches the safety upper limit. Any spatial region below the surface represents a safe region where the motor operating power is below the safety upper limit, while the region above the surface is an overload region. The spatial surface generated by the fitting is defined as the three-dimensional power load constraint boundary.

[0088] Please refer to Figure 5 , step S4 is specifically:

[0089] S401: quantize the hydraulic impedance characteristics of the current pipe network according to the comparison relationship between the water flow of the primary network pipe network and the real-time pressure of the central pipeline in the three-dimensional operating condition data set, fit a trajectory line reflecting the physical state and flow characteristics of the current pipe network, and generate a primary network pipe network resistance characteristic curve;

[0090] According to the primary network pipe network water flow and central pipeline real-time pressure data recorded in the three-dimensional operating condition data set, the fluid mechanics principle is used for reverse deduction. The real-time pressure of the central pipeline is set to approximately represent the pressure drop loss of the pipe network, and the pipe network water flow is the fluid flux. Calculate the ratio of pressure to flow square to get the resistance coefficient. For example, the central pipeline real-time pressure is 0.5 MPa, and the flow is 50 tons per hour. Calculate the resistance coefficient: The average resistance coefficient under the current pipe network state is calculated through regression analysis of multiple groups of pressure and flow data in a recent period of time. Subsequently, a quadratic function relationship between pressure and flow is constructed according to the average resistance coefficient: the head loss is equal to the average resistance coefficient multiplied by the square of the flow. The function relationship is plotted in a three-dimensional space to generate a trajectory line reflecting the current pipe network physical flow capacity and the degree of hindering water flow, that is, the primary network pipe network resistance characteristic curve.

[0091] S402: Set a frequency domain scanning interval covering the rated operation range of the secondary pump, and discretize the frequency domain scanning interval into multiple discrete scanning frequency points using a preset step size. Determine the contact position of the primary network pipe network resistance characteristic curve and the three-dimensional power load constraint boundary at each frequency point in the frequency-flow-power three-dimensional coordinate space one by one, calculate the flow value of the spatial intersection point of the contact position, and lock the physical limit flow at each frequency;

[0092] The process of determining the contact position of the primary network pipe network resistance characteristic curve and the three-dimensional power load constraint boundary at each frequency point in the frequency-flow-power three-dimensional coordinate space is as follows:

[0093] For each discrete scanning frequency point, find the intersection state point in the frequency-flow-power three-dimensional coordinate space that simultaneously satisfies the function constraint of the primary network pipe network resistance characteristic curve and the spatial geometric definition of the three-dimensional power load constraint boundary;

[0094] Analyze the projection value of the intersection state point on the flow coordinate axis to establish the physical limit flow at the corresponding frequency;

[0095] Set a frequency domain scanning interval (e.g., 30 Hz to 50 Hz) covering the rated operation range of the secondary pump, and discretize the interval into multiple discrete scanning frequency points using a preset scanning step size (e.g., 0.1 Hz). In the frequency-flow-power three-dimensional coordinate space, for each discrete scanning frequency point, first determine a slice plane perpendicular to the frequency axis in the space according to the frequency value. Project the primary network pipe network resistance characteristic curve and the three-dimensional power load constraint boundary in the plane simultaneously. Find the intersection point of the two geometric trajectories in the slice plane. The intersection state point means that at the specific frequency, the pump is operating in a state that meets the pipe network resistance demand and just reaches the upper limit of power safety. Analyze the projection value of the intersection state point on the flow coordinate axis, which is the physical limit flow at the frequency. For example, on the 40 Hz slice plane, the pipe network resistance curve and the power boundary surface intersect at a flow of 140 tons per hour, so the physical limit flow at 40 Hz is locked at 140 tons per hour.

[0096] S403: Map the physical limit flow corresponding to all discrete frequency points to the frequency domain analysis plane, logically connect each limit flow point in the order of increasing frequency, and draw the operating range of the secondary pump under the conditions of meeting power safety and pipe network resistance to generate the frequency domain flow reachable interval boundary line;

[0097] The physical limit flow corresponding to all discrete frequency points is extracted to construct a data set containing frequency and limit flow. These data points are mapped to the frequency domain analysis plane with frequency as the horizontal axis and flow as the vertical axis. In the order of increasing frequency, each limit flow point is logically connected to form a continuous boundary line. The boundary line intuitively divides the feasible region and the infeasible region, and the region on one side of the boundary line represents the safe operating range of the secondary pump that can overcome the pipe network resistance to transport water flow and will not cause motor overload. The generated boundary line is defined as the frequency domain flow reachable interval boundary line.

[0098] Please refer to Figure 6 , step S5 is specifically:

[0099] S501: Obtain the preset primary network heating target pressure, combine the real-time pressure of the central pipeline, calculate the numerical difference between the two to establish the optimization direction of the pressure tracking accuracy, and simultaneously retrieve the current operating state coordinates of the secondary pump from the three-dimensional operating condition data set. Calculate the Euclidean distance of the operating state coordinates in the frequency-flow-power three-dimensional space projection plane from the frequency domain flow reachable interval boundary line to establish the optimization direction of the power safety margin, and generate a multi-objective optimization function;

[0100] The process of generating a multi-objective optimization function is specifically:

[0101] A first evaluation dimension is constructed to minimize the absolute value of the difference between the real-time pressure of the central pipeline and the primary network heating target pressure, which measures the accuracy of pressure tracking;

[0102] A second evaluation dimension is constructed to maximize the spatial distance from the current operating state coordinates of the secondary pump to the frequency domain flow reachable interval boundary line, which measures the safety margin of the operating condition;

[0103] The first evaluation dimension and the second evaluation dimension are logically combined to form a multi-objective optimization function that quantitatively evaluates both pressure deviation and safety distance;

[0104] Obtain the preset primary network heating target pressure (e.g., 0.60 MPa). Read the real-time pressure of the central pipeline (e.g., 0.55 MPa). Calculate the absolute value of the difference between the two: , the calculation result is 0.05 MPa, which is used as the first evaluation dimension data for measuring the accuracy of pressure tracking. At the same time, the current running state coordinates (frequency, flow, power) of the secondary pump are retrieved from the real-time data. The coordinates are projected onto the frequency-flow two-dimensional plane, and the Euclidean distance from the projection point to the boundary line of the frequency domain flow reachable interval is calculated. For example, the current point coordinates are (40, 120), and the nearest boundary point coordinates are (42, 130). The Euclidean distance is calculated as follows: . This value 10.2 is used as the second evaluation dimension data for measuring the safety margin of the running condition. A multi-objective optimization function is constructed, and the two indicators are combined using the linear weighting method. The weight coefficients of the first evaluation dimension and the second evaluation dimension are set. The setting of the weight coefficient is based on the following: the fluctuation variance of the complaint rate caused by the pressure deviation in the historical running data, and the fluctuation variance of the equipment alarm caused by the insufficient safety distance, and the normalized ratio of the reciprocal of the two variances (for example, the reciprocal of the pressure deviation variance is 0.6, and the reciprocal of the safety distance variance is 0.4) is calculated as the corresponding weight. The function form is expressed as the target value equal to the first weight coefficient multiplied by the reciprocal of the pressure difference plus the second weight coefficient multiplied by the safety distance. The comprehensive score is calculated as follows: . The result shows the comprehensive score of the current condition, and the larger the value represents the better the comprehensive performance.

[0105] S502: Set the boundary line of the frequency domain flow reachable interval as a hard constraint condition, input the multi-objective optimization function into the NSGA-II multi-objective optimization algorithm, perform multiple rounds of iterative optimization calculation, evaluate the function performance under multiple frequency settings in the global search process, eliminate solutions that do not meet the boundary constraint, locate the best condition configuration, and extract the optimal frequency value;

[0106] The boundary line of the frequency domain flow reachable interval is set as a hard constraint condition. The multi-objective optimization function is input into the multi-objective optimization algorithm. The algorithm initializes a population containing several groups of random frequency setting values. Multiple rounds of iterative optimization calculation are performed, and in each round of iteration, a new generation of population is generated through selection, crossover, mutation and other operations. For each individual, calculate its corresponding multi-objective function value, and judge whether its corresponding working condition point is located within the boundary line of the frequency domain flow reachable interval. Eliminate solutions that do not meet the boundary constraint. After multiple rounds of iteration convergence, the best condition configuration is located, and the optimal frequency value (for example, 42.5 Hz) is extracted.

[0107] S503: Map the optimal frequency value to a standardized signal suitable for the control logic of the frequency converter, analyze the frequency regulation amplitude and direction information of the first-level network secondary pump of the heat station, and generate a frequency control instruction;

[0108] The optimal frequency value is mapped to a standardized signal suitable for the frequency converter control logic. The frequency adjustment magnitude and direction information of the heat station primary network secondary pump is parsed. For example, the current frequency is 40 Hz, and the optimal frequency is 42.5 Hz, so the adjustment direction is to increase, and the adjustment magnitude is 2.5 Hz. The optimal frequency value is converted into a control instruction recognizable by the frequency converter, such as a Modbus register hexadecimal value or a 4-20 mA analog signal. For example, using the standard of 0-50 Hz corresponding to 4-20 mA, the control signal current value is calculated as follows: The calculation result is 17.6 mA. Based on the calculation result, a control frequency instruction containing a 17.6 mA current set value is generated, and the instruction is sent to the secondary pump frequency converter through the industrial field bus to drive the water pump motor to adjust the operating frequency to the optimal state.

[0109] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A multi-objective optimized method for the operation and control of secondary pumps in a primary network of a thermal power station, characterized in that, Includes the following steps: S1: Construct the hydraulic connectivity topology of the primary network based on the physical connection sequence of the supply and return water pipelines and the valve opening and closing status of the primary network of the heating station, divide the hydraulic control zones, and collect the zone pipeline pressure data set for each hydraulic control zone. S2: Based on the set of zoned pipeline pressure data, filter the hydraulic feature center pipelines and extract the real-time pressure of the center pipelines; S3: Synchronously acquire the operating frequency of the secondary pump, the water flow rate of the primary network, and the real-time power of the secondary pump motor at the current moment, construct a three-dimensional operating condition dataset, and divide the three-dimensional power load constraint boundary through preset safety judgment logic; S4: Based on the relationship between the current primary network water flow and the real-time pressure of the central pipeline, construct the primary network resistance characteristic curve, scan the intersection of the curve and the three-dimensional power load constraint boundary, and construct the frequency domain flow reachable interval boundary line; S5: The first optimization objective is to minimize the deviation between the real-time pressure of the central pipeline and the target pressure of the primary network heating, and the second optimization objective is to generate a frequency control command based on the current operating point of the secondary pump and the boundary line of the frequency domain flow reachable interval. Step S3 is as follows: S301: Synchronously collect three data points of the secondary pump at the current moment: the rated frequency range of the secondary pump, the operating frequency of the secondary pump, the water flow of the primary network, and the real-time power of the secondary pump motor. Establish a three-dimensional coordinate space with frequency, flow, and power as three dimensions, map the three collected data points to coordinate points in the space, and generate a three-dimensional operating condition dataset. S302: Compare the real-time power of the secondary pump motor in the three-dimensional operating condition dataset with the preset rated power benchmark, remove abnormal data that exceed the power limit, traverse and retrieve the flow record at each frequency point within the rated frequency range of the secondary pump, extract the maximum primary network water flow value achieved under the power safety limit, and generate the non-overload maximum flow node. S303: Remap the non-overload maximum flow node back to the frequency-flow-power three-dimensional coordinate space, and connect the maximum flow and corresponding power data at each frequency point to form a continuous spatial surface or geometric envelope, establish the power safety range of the secondary pump under multiple operating frequency and flow combinations, and generate a three-dimensional power load constraint boundary. Step S4 is as follows: S401: Based on the comparison between the water flow rate of the primary network and the real-time pressure of the central pipeline in the three-dimensional operating condition data set, the hydraulic impedance characteristics of the current network are quantified and fitted into a trajectory line that reflects the physical state and flow characteristics of the current network, thereby generating the resistance characteristic curve of the primary network. S402: Set a frequency domain scanning interval covering the rated operating range of the secondary pump, and use a preset step size to discretize the frequency domain scanning interval into multiple discrete scanning frequency points. In the frequency-flow-power three-dimensional coordinate space, determine the contact position between the primary network resistance characteristic curve and the three-dimensional power load constraint boundary at each frequency point, calculate the flow rate value at the spatial intersection of the contact position, and lock the physical limit flow rate at each frequency. S403: Map the physical limit flow rates corresponding to all discrete frequency points to the frequency domain analysis plane, logically connect each limit flow rate point in ascending order of frequency, define the operating range of the secondary pump under the conditions of power safety and pipeline resistance, and generate the boundary line of the frequency domain flow rate reachable interval.

2. The multi-objective optimized operation and control method for the secondary pumps of a primary network in a thermal power station according to claim 1, characterized in that, The partitioned pipeline pressure data set includes hydraulic control partition codes based on physical connections, supply and return water pipeline location indices mapped to each partition, and real-time pressure monitoring values ​​corresponding to each index. The real-time pressure of the central pipeline includes hydraulic feature central pipeline identifiers filtered based on consistency offsets and the current real-time pressure measurement readings of the pipeline. The three-dimensional power load constraint boundary specifically refers to a discrete frequency sampling sequence that meets the rated power conditions and the maximum primary network water flow extreme value corresponding to each sampling frequency point. The frequency domain flow reachable interval boundary line specifically refers to the spatial intersection coordinates of the three-dimensional power load constraint boundary and the primary network resistance characteristic curve, and the flow physical limit contour formed based on the intersection connection. The frequency control command includes a variable frequency drive target operating frequency setting command and a speed adjustment execution command to drive the secondary pump.

3. The multi-objective optimized operation and control method for the secondary pumps of a primary network in a thermal power station according to claim 1, characterized in that, Step S1 is as follows: S101: Obtain the physical connection sequence of the supply and return water pipelines and the valve opening and closing status of the primary network of the heating station, establish the pipeline node connection relationship based on the physical connection sequence, determine the conduction status between pipelines in combination with the valve opening and closing status, and generate the hydraulic connectivity topology of the primary network. S102: Retrieve the current status of all valves in the primary network hydraulic connectivity topology, identify valves in the closed state and mark them as physical isolation boundaries, decompose the overall connectivity topology into independent subnets that are not connected to each other based on the physical isolation boundaries, determine the pipeline range covered by each independent subnet, and generate independent hydraulic control zones; S103: Use pressure sensors located at each pipeline position to collect the real-time pressure of the primary network supply and return water pipelines at the same time, obtain the physical location information of the pressure sensors, match and map the collected pressure values ​​to the corresponding independent hydraulic control zone range according to the physical location information, and generate a zone pipeline pressure data set.

4. The multi-objective optimized operation and control method for the secondary pumps of a primary network in a thermal power station according to claim 1, characterized in that, Step S2 is as follows: S201: Traverse all real-time pressure data of primary network supply and return water pipelines in the partitioned pipeline pressure data set, and calculate the regional average pressure value of the overall pressure level in the current hydraulic control partition. S202: Extract the real-time pressure of each pipe location in the partitioned pipe pressure data set one by one, calculate the comparison data representing the magnitude of the difference between the real-time pressure of each pipe location and the average pressure value of the region, quantify the degree of deviation of each single-point pressure data relative to the average level of the region, and generate pressure consistency offset. S203: Arrange all the pressure consistency offsets in ascending order of value, lock the smallest offset value at the top, trace the physical location of the primary network pipeline corresponding to the smallest offset data, mark it as the hydraulic feature center pipeline representing the overall hydraulic state of the partition, and extract the real-time pressure of the center pipeline.

5. The multi-objective optimized operation and control method for the secondary pumps of a primary heat station network according to claim 1, characterized in that, The process of establishing a three-dimensional frequency-flow-power coordinate space with three dimensions of frequency, flow rate, and power is as follows: Construct a spatial rectangular coordinate system with the operating frequency of the secondary pump, the water flow rate of the primary network, and the real-time power of the secondary pump motor as independent coordinate axes; The three data items acquired synchronously at the same time are associated and bound together, and defined as a spatial feature vector describing the state at a single running moment; Based on the values ​​of each component of the spatial feature vector, a unique geometric coordinate point is located in the spatial rectangular coordinate system, and all geometric coordinate points in the continuous time series are collected and stored.

6. The multi-objective optimized operation and control method for the secondary pumps of a primary network in a thermal power station according to claim 1, characterized in that, The process of determining the contact position between the next-level network resistance characteristic curve and the three-dimensional power load constraint boundary at each frequency point in the frequency-flow-power three-dimensional coordinate space is as follows: For each discrete scanning frequency point, in the frequency-flow-power three-dimensional coordinate space, find the intersection point of the function constraint of the primary network resistance characteristic curve and the spatial geometric definition of the three-dimensional power load constraint boundary. The projection values ​​of the intersection points onto the flow rate coordinate axis are analyzed and established as the physical limit flow rates at the corresponding frequencies.

7. The method for multi-objective optimized operation and control of secondary pumps in a primary heat station network according to claim 1, characterized in that, Step S5 is as follows: S501: Obtain the preset primary network heating target pressure, combine it with the real-time pressure of the central pipeline, calculate the numerical difference between the two to establish the optimization direction of pressure tracking accuracy, and at the same time retrieve the current operating status coordinates of the secondary pump from the three-dimensional operating condition dataset, calculate the Euclidean distance between the operating status coordinates and the boundary line of the frequency domain flow reachable interval in the frequency-flow-power three-dimensional spatial projection plane, establish the optimization direction of power safety margin, and generate a multi-objective optimization function. S502: Set the boundary line of the reachable interval of the frequency domain flow as a hard constraint condition, input the multi-objective optimization function into the NSGA-II multi-objective optimization algorithm, perform multiple rounds of iterative optimization calculation, evaluate the function performance under various frequency settings during the global search process, eliminate solutions that do not meet the boundary constraints, locate the best working condition configuration, and extract the optimal frequency value; S503: The optimal frequency value is mapped and converted into a standardized signal suitable for the inverter control logic, the frequency adjustment amplitude and direction information of the secondary pump of the primary network of the heating station is analyzed, and the control frequency command is generated.

8. The multi-objective optimized operation and control method for the secondary pumps of a primary heat station network according to claim 7, characterized in that, The process of generating a multi-objective optimization function is as follows: A first evaluation dimension is constructed, guided by minimizing the absolute value of the difference between the real-time pressure of the central pipeline and the target pressure of the primary network heating, to measure the accuracy of pressure tracking; A second evaluation dimension is constructed, guided by maximizing the spatial distance from the current operating state coordinates of the secondary pump to the boundary line of the frequency domain flow reachable interval, to measure the safety margin of the operating condition; By logically combining the first and second evaluation dimensions, a multi-objective optimization function is formed that simultaneously quantifies and evaluates pressure deviation and safety distance.

Images