Complex section open channel flow velocity comprehensive solving method under multi-point emission mechanism

Abstract

The invention relates to the technical field of numerical simulation modeling, in particular to a complex section open channel flow velocity comprehensive solving method under a multi-point emission mechanism, which comprises the following steps of: performing joint grouping, density statistics and sequence screening on open channel section parameters, disturbance source distribution characteristics and water level gradient data; a dynamic partitioning process driven by structural parameters is established, a lattice structure better fitting non-uniform boundary characteristics is constructed, extension layering is performed on a disturbance projection direction and a propagation path, and a graph neural network is fused to extract the correlation degree between disturbance nodes and construct a propagation path graph, so that the propagation path graph is extracted, and the propagation path graph is constructed. The problem of local propulsion offset in particle propulsion is effectively eliminated, continuity of multi-point disturbance response in an open channel and coordination of particle propagation paths are guaranteed by gradually constructing a multi-scale disturbance response mechanism and dynamic synchronous propulsion logic, and the distribution simulation capability of irregular boundaries and dynamic disturbance fields is remarkably enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of numerical simulation modeling, and in particular to a complex cross-section open channel flow velocity comprehensive solution method under a multi-point emission mechanism. BACKGROUND

[0002] The technical field of numerical simulation modeling mainly involves modeling and simulation of natural systems using computers, and the goal is to accurately describe physical processes and discretize them into mathematical models for simulation and solution on a digital platform. The field of numerical simulation modeling usually combines partial differential equation modeling theory, finite element numerical method, mesh generation technology and high-performance computing resources to solve complex boundary and multi-variable coupling problems in natural systems.

[0003] A complex cross-section open channel flow velocity comprehensive solution method under a multi-point emission mechanism, by constructing a multi-point emission mechanism, selecting multiple emission points in the calculation domain as initial disturbance sources and boundary driving, realizing comprehensive description of complex water flow motion behavior, the purpose is to establish a flow velocity field model suitable for irregular boundary conditions, to improve the accuracy of open channel flow velocity distribution prediction, by introducing multi-point disturbance and joint solution, to overcome the problem of insufficient response to local flow velocity abnormal points under single-point boundary driving, to realize continuous, stable and high-resolution simulation of the entire cross-section flow velocity state.

[0004] The existing technology focuses on disturbance modeling based on single-level boundary conditions and uniform emission points, which is prone to problems such as response blind area and synchronous jump identification failure, lacks differentiated response processing of disturbance path dynamic evolution and particle propulsion process, resulting in particle accumulation, propulsion sequence misplacement and jump peak identification failure in complex cross-sections, and due to the lack of parallel regulation and control of time difference and spatial expansion in the propulsion path, local flow velocity value is often abnormally high and underestimated, affecting the overall flow field solution accuracy, causing the simulation structure to lose consistency, and reducing the overall simulation robustness and application range. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art and to provide a complex cross-section open channel flow velocity comprehensive solution method under a multi-point emission mechanism.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme: a complex cross-section open channel flow velocity comprehensive solution method under a multi-point emission mechanism, comprising the following steps: Step 1: Based on the open channel cross-section structure parameters, the number of disturbance sources and the slope fluctuation coefficient, calculate the ratio of cross-section coordinates to area, group the difference between disturbance source spacing and flow inclination, and count the water level gradient and flow velocity intersection point density to generate a flow basin dynamic zoning structure delineation map; Step two: based on the dynamic zoning structure delineation map of the watershed, the main stream segment velocity and water level sequence are staggered and compared, the boundary segment displacement and the angle between the velocity are screened, the boundary slope angle and the depth jump point are sorted, and a difference grid driving structure set is generated; Step three: based on the difference grid driving structure set, the disturbance source coordinates and the projection direction are extended, the correlation between the nodes of the disturbance path is analyzed by using a graph neural network, and a mutation propagation structure graph is constructed, the disturbance amplitude and the propagation change rate are sorted in layers, and a disturbance action area grid synchronization instruction table is generated; Step four: based on the disturbance action area grid synchronization instruction table, the particle propagation time sequence and the advance difference are obtained, the overlap boundary advance time difference matrix is sorted, the advance direction density block is matched and numbered, the particle advance sequence parallel computing scheduling table is generated; Step five: based on the particle advance sequence parallel computing scheduling table, the boundary segment state transition and the advance period difference are extracted, the advance order of the mutation area is rearranged, the overlap boundary synchronous advance section is reconstructed, and a multi-region grid state coupling advance structure graph is generated.

[0007] As a further scheme of the present application, the watershed dynamic zoning structure delineation map includes zoning boundary lines, disturbance response area labels and hydraulic slope change area identification codes, the difference grid driving structure set includes grid classification identification, boundary difference area number and jump correlation segment index, the disturbance action area grid synchronization instruction table includes node matching path number, synchronous advance level code and disturbance response grid index, the particle advance sequence parallel computing scheduling table includes advance task segmentation sequence, parallel sub-block number and scheduling time step mapping table, and the multi-region grid state coupling advance structure graph includes grid boundary connection graph, coupling advance channel mapping and synchronous state marker field.

[0008] As a further scheme of the present application, the specific steps for generating the watershed dynamic zoning structure delineation map are: Based on the open channel cross-section structure parameters, the number of disturbance sources and the slope fluctuation coefficient, the sectional distance between the cross-section coordinate point groups is determined and the contour closed curve area is derived, the disturbance source interconnection length valley value and the adjacent water flow direction angle difference value are extracted and classified, and the cross-section and disturbance relationship extraction result is generated; Based on the cross-section and disturbance relationship extraction result, the elevation difference between the water level sampling points in the longitudinal slope area and the gradient ratio are counted, the density distribution of the intersection points of the transverse velocity vectors is identified, and the overlapping area of the turning cross-section position is screened, and a watershed dynamic zoning structure delineation map is generated.

[0009] As a further scheme of the present application, the specific steps for generating the difference grid driving structure set are: Based on the basin dynamic zoning structure delineation map, the water level record value and the corresponding time point speed record value of each cross section position in the main flow section are aligned, the difference value of the water level and the speed of the same position in the adjacent time step is subtracted, the trend comparison is calculated, and the matching point sequence is labeled, and a slope change contrast set of the main flow section is generated; Based on the slope change contrast set of the main flow section, the included angle between the trajectory line of each control point in the boundary section and the angle relationship of the adjacent velocity vector is extracted, the control points with an included angle less than a specified angle range are obtained, the slope curvature change value of the control point is extracted, and the longitudinal depth difference sequence is sequentially sorted, and a difference grid driving structure set is generated.

[0010] As a further scheme of the present application, the specific steps for generating the disturbance action area grid synchronization instruction table are: Based on the difference grid driving structure set, the disturbance source coordinate point and the number of the starting line segment direction of the grid boundary where the disturbance source coordinate point is located are bound, and a line segment with a fixed step length is added in the diagonal direction from the bound point, the displacement node coordinates are recorded, the cumulative length and the end point of the line segment are calculated and labeled, and a disturbance source extension path data set is generated; Based on the disturbance source extension path data set, the local speed data of each extension node and the disturbance frequency change value of the adjacent node are subtracted to generate a difference sequence, a graph neural network is used to extract the disturbance correlation between nodes and construct a jump propagation path graph, assist the frequency transfer trajectory between the matching mutation threshold nodes, and position the nodes with a difference value greater than the jump threshold value in each segment, bind the direction tracking label and the corresponding path identification code, and generate a disturbance mutation node matching set; Based on the disturbance mutation node matching set, the disturbance amplitude peak value corresponding to each node and the propagation speed difference of the nodes in the downstream propagation path are extracted, the amplitude value in each segment is sorted, the speed difference is divided, the flag bit is updated synchronously, and the target grid number is reorganized to generate a disturbance action area grid synchronization instruction table.

[0011] As a further scheme of the present application, the graph neural network extracts the disturbance correlation between nodes and constructs a jump propagation path graph, first, the extension node is regarded as an independent node in the graph structure, a directed edge connection structure is established according to the spatial adjacency relationship between nodes and the frequency change sequence, the disturbance frequency difference between nodes and the spatial displacement are used as the basis for edge weight assignment, after the graph structure is generated, the disturbance correlation index between nodes is iteratively calculated, the strong node pairs in the disturbance propagation chain are extracted, and a frequency jump path index graph is constructed.

[0012] As a further scheme of the present application, the diagonal direction additional fixed step progression line segment specifically refers to taking the starting line segment direction of the lattice boundary where the disturbance source is located as the reference direction, determining the diagonal extension direction with the relative peak angle offset in the lattice structure, and extending outward in the direction, simultaneously increasing the same value as the fixed step in the horizontal direction and the vertical direction in the two-dimensional coordinate system, drawing a line segment consistent with the initial direction at each step position, with the starting point of the line segment being the end point of the previous node and the end point being the current step coordinate point, recursively until the region boundary is reached and the preset distance condition is met, and the starting and ending coordinates of the line segment are recorded, and the total length and end point information are updated based on the length of each step.

[0013] As a further scheme of the present application, the specific steps for generating the particle advance sequence parallel computing scheduling table are: Based on the disturbance action area lattice synchronization instruction table, the lattice numbers of each synchronization entry in the table are extracted and the time field is read, the number is extracted by reading the instruction line, a time stamp is generated, and a time label is performed, the particle advance sequence difference value and the node time difference are calculated, and a particle advance time sequence difference list is generated; Based on the particle advance time sequence difference list, the particle set corresponding to the time difference entry and the boundary unit number are matched, the particle ID and unit number mapping are read, and the time difference aggregation and intersection area in the group are identified, the time difference matrix row and column indexes are sorted, and a boundary advance difference sorting matrix set is generated. Based on the boundary advance difference sorting matrix set, the particle number frequency of occurrence at the intersection point of the matrix is counted, the frequency continuous segment is identified by counting the number list of each intersection point, the block number allocation and the advance sequence are adjusted, and a particle advance sequence parallel computing scheduling table is generated.

[0014] As a further scheme of the present application, the particle advance sequence difference value refers to the sequence difference between the time labels corresponding to different particles on their respective advance paths in the particle advance scheduling, specifically based on the time field read by the particle in the synchronization instruction table, the particle advance time label is generated, and then the particles are sorted according to the lattice number and the advance path, and the difference between the advance label values of adjacent particles is calculated.

[0015] As a further scheme of the present application, the specific steps for generating the multi-region lattice state coupling advance structure diagram are: Based on the particle advance sequence parallel computing scheduling table, the continuity of the particle state label and the advance period marker field in each boundary segment is read, the period sequence between the advance start node and the end node is extracted and compared piece by piece, the peak and valley difference between the periods is identified and the advance offset sorting is performed, and a boundary segment advance difference data set is generated. Based on the boundary segment, the advancing difference data set is promoted, the particle number order in the advancing offset segment is rearranged and combined with the time period index value, the node number of the continuous mutation area is located and the synchronous advancing number list is extracted, the advancing segment is recombined, the boundary coupling area is combined with the synchronous grid block path, and the multi-area grid state coupling advancing structure diagram is generated.

[0016] Compared with the prior art, the advantages and positive effects of the present application are that: 1. In the present application, by jointly grouping, density statistics and sequence screening of the open channel section parameters, disturbance source distribution characteristics and water level gradient data, a dynamic partition process driven by structural parameters is established, and a grid structure more suitable for non-uniform boundary characteristics is constructed, which significantly improves the overall coordination accuracy of the flow velocity calculation field and the calculation efficiency of the grid system; 2. In the present application, by extending and layering the disturbance projection direction and the propagation path, and by fusing the graph neural network to extract the correlation between disturbance nodes and construct the propagation path graph, the path recognition process is more accurate and dynamically adaptable, effectively eliminating the local advancing offset problem in particle advancing; 3. In the present application, by gradually constructing a multi-scale disturbance response mechanism and a dynamic synchronous advancing logic, the continuity of multi-point disturbance response and the coordination of particle propagation path in the open channel are guaranteed, and the distribution modeling capability for irregular boundaries and dynamic disturbance fields is significantly enhanced. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 The present application is a schematic diagram of the main steps. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0019] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only used to facilitate the description of the present application and simplify the description, and therefore cannot be understood as indicating or implying that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0020] EMBODIMENT Please refer to Figure 1The application provides a technical scheme: a complex cross-section open channel flow velocity comprehensive solution method under a multi-point transmitter mechanism, which comprises the following steps: Step one: based on the open channel cross-section structure parameters, the number of disturbance sources and the slope fluctuation coefficient, the ratio of the cross-section coordinate to the area is calculated, the distance between the disturbance sources and the flow inclination difference is grouped, the water level gradient and the flow velocity intersection point density are counted, and a flow basin dynamic partition structure demarcation map is generated; Step two: based on the flow basin dynamic partition structure demarcation map, the main stream velocity and water level sequence are compared, the boundary segment displacement and the velocity angle are screened, and the boundary slope angle and the depth jump point are sorted, and a difference grid driving structure set is generated; Step three: based on the difference grid driving structure set, the disturbance source coordinates and the projection direction are extended, the correlation between the nodes on the disturbance path is analyzed by using a graph neural network, and a mutation propagation structure diagram is constructed, the disturbance amplitude and the propagation change rate are layered and sorted, and a disturbance action area grid synchronous instruction table is generated; Step four: based on the disturbance action area grid synchronous instruction table, the particle propagation time sequence and the promotion difference are obtained, the overlap boundary promotion time difference matrix is sorted, the promotion direction density block is matched and numbered, the particle promotion sequence parallel computing scheduling table is generated; Step five: based on the particle promotion sequence parallel computing scheduling table, the boundary segment state transition and the promotion period difference are extracted, the promotion order of the mutation area is rearranged, the overlap boundary synchronous promotion section is reconstructed, and a multi-region grid state coupling promotion structure diagram is generated.

[0021] The flow basin dynamic partition structure demarcation map comprises partition boundary lines, disturbance response area labels and hydraulic slope change area identification codes, the difference grid driving structure set comprises grid classification identification, boundary difference area number and jump correlation segment index, the disturbance action area grid synchronous instruction table comprises node matching path number, synchronous promotion level code and disturbance response grid index, the particle promotion sequence parallel computing scheduling table comprises promotion task segmentation sequence, parallel sub-block number and scheduling time step mapping table, and the multi-region grid state coupling promotion structure diagram comprises grid boundary connection map, coupling promotion channel mapping and synchronous state marker field.

[0022] The specific steps for generating the flow basin dynamic partition structure demarcation map are as follows: Based on the open channel cross-section structure parameters, the number of disturbance sources and the slope fluctuation coefficient, the sectional distance between the cross-section coordinate point groups is determined, and the profile closed curve area is derived, the disturbance source interconnection length valley value and the adjacent water flow direction angle difference value are extracted, and the interval classification is performed, and the cross-section and disturbance relationship extraction result is generated; Based on the cross-section and disturbance relationship extraction result, the elevation difference and gradient ratio between the water level sampling points in the longitudinal slope area are counted, the density distribution of the horizontal velocity vector intersection points is identified, and the turning cross-section position overlap area is screened, and the flow basin dynamic partition structure demarcation map is generated. Based on the open channel cross-section structure parameters, the number of disturbance sources and the slope fluctuation coefficient, the cross-section polygon node set is obtained using complex coordinate representation, the nodes are sequentially numbered, the contour closing curve calculation is performed according to Green's formula, the integral starting point is set as the lowest numbered point, and the clockwise direction is executed, the area is calculated as one-half times the sum of the product difference of the horizontal and vertical coordinates of the nodes, and then the boundary envelope is limited combined with the elevation difference, and the cross-section and disturbance relationship extraction result is generated; Based on the cross-section and disturbance relationship extraction result, a two-dimensional elevation vector array is constructed for the sampling points in the longitudinal slope interval, a two-dimensional vector array composed of elevation difference and gradient ratio is clustered, the K value is set to 5, the initial center point is constructed from the maximum gradient position, the attribution of each type of sample is completed, the overlapping section is screened, then a two-dimensional density analysis method is used, a planar distribution density map of the intersection point of the horizontal velocity vector is constructed using kernel density estimation, the kernel function is set to Gaussian kernel, the bandwidth parameter is set to 0.8, the coordinate points exceeding the density threshold are screened for turning cross-section overlapping positions, and a dynamic partition structure delineation map of the basin is generated.

[0023] The specific steps for generating the difference grid driving structure set are as follows: Based on the dynamic partition structure delineation map of the basin, the water level record value and the corresponding time point velocity record value of each cross-section position in the main flow section are aligned, and the difference between the water level and the velocity of the same position in the adjacent time step is subtracted, the trend comparison is calculated, and the matching point sequence is labeled to generate the main flow section slope change comparison set; Based on the main flow section slope change comparison set, the angle between the trajectory line of each control point in the boundary section and the adjacent velocity vector is extracted, the control points with an included angle less than a specified angle range are obtained, the slope curvature change value of the control point is extracted, and the longitudinal depth difference sequence is sequentially sorted to generate the difference grid driving structure set; Based on the dynamic partition structure delineation map of the basin, a two-channel time series array containing water level and velocity values is constructed for each cross-section position in the main flow section, the water level and velocity sequence are aligned, the difference of the same cross-section position in the adjacent time step is calculated, the time interval is set to 5 seconds, the data accuracy is retained to two decimal places, each group of difference values is calculated and a trend difference vector is constructed, then a dynamic time warping algorithm is used to construct a mapping for the matching points in the trend difference vector, a search path range with a constraint window of 6 is used, and the matching point position is written into a matching array, finally a one-dimensional label sequence is constructed by assigning a unique label value to each point in the matching array, and a main flow section slope change comparison set is generated. Based on the slope change of the mainstream section, the control point reading trajectory line coordinate sequence of the boundary section is calculated, and the tangent derivative is calculated. The derivative is multiplied by the velocity vector of the adjacent node, and the inverse cosine of the modulus product is calculated. The angle is set to 30 degrees, only the control point index less than the threshold value is retained, and the trajectory line of the retained point is executed by the curvature difference calculation method. The second difference operation is constructed to build a curvature sequence. After normalization using the maximum value, the curvature is sorted, and the sorted index and depth difference value are merged in coordinate order. The joint sorting is executed. The priority rule is that the curvature order is in front, and the depth order is in back. The difference grid driving structure set is generated.

[0024] The specific steps for generating the perturbation action area grid synchronization instruction table are as follows: Based on the difference grid driving structure set, the perturbation source coordinate point is bound with the number of the starting line segment direction of the grid boundary, and the line segment is added in the diagonal direction with a fixed step length. The displacement node coordinates are recorded, the cumulative length and the end point of the line segment are calculated and marked, and the perturbation source extension path data set is generated. Based on the perturbation source extension path data set, the local speed data of each extension node and the adjacent node perturbation frequency change value are subtracted to generate a difference sequence. The graph neural network is used to extract the perturbation correlation between nodes and construct a jump propagation path graph. The frequency transfer trajectory between the nodes matched with the mutation threshold is assisted. The nodes with a difference value greater than the jump threshold in each segment are paired in position, and the direction tracking mark is bound with the corresponding path identification code to generate a perturbation mutation node matching set. Based on the perturbation mutation node matching set, the perturbation amplitude peak value corresponding to each node and the propagation speed difference of the nodes in the downstream propagation path are extracted, and the amplitude value in each segment is sorted. The speed difference is differentiated, the flag bit is updated synchronously, and the target grid number is reorganized to generate a perturbation action area grid synchronization instruction table. Based on the difference grid driving structure set, the starting line segment of the grid boundary where the perturbation source coordinate point is located is positioned. The line segment direction number 0 to 7 corresponds to the main diagonal and the auxiliary diagonal direction. After binding the number, the directional addition is performed according to the specified step length value of 3 units. The end points of the subsequent line segments are generated in turn using the incremental iteration method. The coordinates of each displacement node are recorded and the cumulative length value is updated after generation. The termination condition is that the cumulative length is greater than or equal to 15 units and the boundary section is encountered. The end point coordinates and the cumulative length information are written into the path array at the same time to generate the perturbation source extension path data set. Based on the disturbance source extended path dataset, a graph neural network method is used to construct a node graph structure. The local velocity data of each extended node and the disturbance frequency change value of the adjacent node are used as the input feature tensor. The adjacency matrix is ​​constructed to connect the upstream and downstream nodes in the path. The GCNConv layer is defined with the number of layers set to 2, the activation function is ReLU, the number of graph network training rounds is set to 300 rounds, the learning rate is 0.005, and the loss function is the mean square error function. Forward propagation is performed on the input graph data, and the disturbance correlation matrix is ​​output. The jump connection edges are filtered according to the maximum response value between nodes, and a frequency jump propagation path graph is constructed. The nodes in each segment difference sequence that are greater than the specified jump threshold of 0.6 are sequentially paired according to the propagation path graph. The direction vectors of the paired nodes are marked and the path identification code is attached to generate a disturbance mutation node matching set. Based on the perturbation mutation node matching set, the amplitude peak of each mutation node within the third-order time step is extracted, and the local maximum function with a sliding window length of 3 is used to identify the peak index of each segment. Then, the propagation speed is calculated node by node in the downstream path. The speed difference is the speed of the adjacent node minus the speed of the previous node. The amplitude peaks in the same path are sorted by size, and the standard deviation of the difference is classified. The interval boundaries are set to 0.2, 0.5 and 0.8, and hierarchical indexing is performed. The node array is labeled and the flag field is updated according to the level. Then, the corresponding grid number is written synchronously to generate the grid synchronization instruction table of the perturbation action area.

[0025] The graph neural network extracts the perturbation correlation between nodes and constructs a frequency hopping propagation path graph. First, the extended nodes are regarded as independent nodes in the graph structure. A directed edge connection structure is established based on the spatial adjacency relationship and frequency change sequence between nodes. The perturbation frequency difference and spatial displacement between nodes are used as the basis for edge weight assignment. After the graph structure is generated, the perturbation correlation index between each node is iteratively calculated to extract the highly correlated node pairs in the perturbation propagation chain and construct a frequency hopping path index graph. The graph neural network follows the formula: in: Representing the graph neural network Layer The embedding vector update result of each node is: Representing the graph neural network Layer The embedding of adjacent nodes, Indicates the The set of adjacent nodes of a node, Representing the graph neural network The weight matrix of the layer represents the activation function, Indicates the Node and The disturbance speed difference factor between adjacent nodes, Indicates the Node and The frequency jump significance coefficient between adjacent nodes, Indicates the Node and The path direction consistency factor between adjacent nodes, Indicates the Node and The normalization constant between adjacent nodes.

[0026] Execution process: First, the open channel section is discretized into multiple disturbance source node sets, and the embedded representation of each node at the current moment is extracted. , including speed, frequency, direction and other attribute values, traverse the nodes Adjacent nodes The speed difference factor is calculated by normalizing the propagation speed difference between the two nodes by the maximum speed difference. The frequency jump factor is calculated by dividing the standard deviation of the disturbance frequency change by the specified jump threshold , the path direction consistency factor is calculated by the cosine value of the angle between the disturbance direction vector and the mainstream direction , take the product of the three as the weight of the adjacent edge and divide it by the normalization term Balance the influence of adjacent structure density, and multiply each weight by the adjacent node characteristics of the previous layer Then sum and enter the activation function , complete the Layer node representation After multiple layers of propagation, a jump propagation path graph is formed.

[0027] Adding a line segment with a fixed step length in the diagonal direction specifically means taking the starting line segment direction of the grid boundary where the disturbance source is located as the reference direction, determining the diagonal extension direction with a relative peak angle offset in the grid structure, and extending outward in the direction, and increasing the same value in the horizontal and vertical directions at the same time as the fixed step length in the two-dimensional coordinate system, drawing a line segment consistent with the initial direction at each step position, the starting point of the line segment is the end point of the previous node, and the end point is the current step coordinate point, and recursively in sequence until it touches the boundary of the area and meets the preset distance condition. The starting and ending coordinates of the line segment need to be recorded, and the total length and end point information are updated based on the cumulative length of each step.

[0028] The specific steps to generate the particle advancement sequence parallel computing schedule are: Based on the lattice synchronization instruction table of the disturbance action area, the lattice numbers of each synchronization entry in the table are extracted and the time field is read. The number is extracted by reading the instruction line, and the timestamp is generated and time-labeled. The particle advance sequence difference value and the node time difference are calculated, and the particle advance time sequence difference list is generated; Based on the particle advance time sequence difference list, the particle set corresponding to the time difference entry and the boundary cell number are matched, the particle ID and cell number mapping are read, and the time difference aggregation and intersection area in the group are identified. The row and column indexes of the time difference matrix are sorted, and the boundary advance difference sorting matrix set is generated. Based on the boundary advance difference sorting matrix set, the frequency of the particle number appearing at the intersection point of the matrix is counted, and the frequency continuous segment is identified by counting the number list of each intersection point. The block number allocation and the advance sequence are adjusted, the particle advance sequence parallel computing scheduling table is generated; Based on the lattice synchronization instruction table of the disturbance action area, the lattice numbers corresponding to each synchronization instruction and the time record values are extracted in sequence. The time record is uniformly converted into a standard format time label, and the advance time comparison between the continuous numbers is carried out after sorting by number. The time difference value of each pair of continuous lattices is calculated, combined with the corresponding particle number identifier, the time difference of each node in the particle advance path is arranged, and the output is the particle advance time sequence difference list. Based on the particle advance time sequence difference list, the mapping relationship between the particle number and the advance difference value is analyzed, the boundary cell number corresponding to the particle set is extracted, the time difference value change amplitude and aggregation degree of multiple particles in the same cell are counted, the area with concentrated time difference is recorded, and whether the advance rhythm of multiple particles in a certain area overlaps is analyzed. The row and column numbers of each particle path are sorted according to the advance difference value, and the output is the boundary advance difference sorting matrix set. Based on the boundary advance difference sorting matrix set, the total number of times the particle number corresponding to the intersection position of the matrix appears is counted, the numbers are segmented in sequence according to the intersection frequency, the number region with time advance continuity and repeatability in the particle group is identified, the region is divided into independent advance block units, and the original particle advance sequence is structurally adjusted. According to the arrangement of the position and the advance rhythm of the particle, the particle advance sequence parallel computing scheduling table is constructed and output.

[0029] The particle advance sequence difference value refers to the sequence difference between the time labels corresponding to different particles in their respective advance paths in the particle advance scheduling. Specifically, based on the time field read by the particle in the synchronization instruction table, the particle advance time label is generated, and then the particles are sorted according to the lattice number and the advance path. The difference between the advance label values of adjacent particles is calculated.

[0030] The specific steps of generating a multi-region lattice state coupling advance structure diagram are as follows: Based on the parallel calculation schedule of the particle advancement sequence, the particle state labels and advancement cycle mark fields in each group of boundary segments are read continuously. The period sequence between the advancement start node and the end node is extracted and the difference range is compared one by one. The peak and valley value differences between the cycles are identified and the advancement offset is sorted to generate the boundary segment advancement difference dataset. Based on the boundary segment propulsion difference dataset, the particle numbering order and time period index value combination in the propulsion offset segment are rearranged, the node numbers of the continuous mutation zone are located and the synchronous propulsion number list is extracted. The propulsion segment reorganization is combined with the synchronous grid block path of the boundary coupling zone to generate a multi-zone grid state coupled propulsion structure diagram; Based on the parallel calculation schedule of the particle advancement sequence, a linear cursor is used to read the particle number sequence and the advancement cycle mark field according to the advancement time index. The starting bit of the field is set as the advancement starting point index, and a sliding window interval with an interval step of 1 is set during the reading process to perform a continuity check on the period value. The advancement start and end nodes are screened out and the intermediate sequence data is obtained. The mean and range of each advancement cycle are compared, and the peak and valley recognition threshold value difference is set between 0.25 and 0.45 as the jump judgment standard. The matching condition nodes in the advancement difference sequence within the cycle segment are classified into secondary intervals and index rearranged. The corresponding relationship between the mark field and the difference table is sorted and numbered, and a table structure is constructed for output to generate a boundary segment advancement difference dataset. Based on the boundary fragment advancement difference dataset, the particle number field and the advancement time field are set as the joint primary key. A particle index cache is constructed within the advancement offset segment and the row and column matrix position numbers corresponding to the time period are read. The particles are arranged in reverse order from high to low advancement differences according to the advancement trend direction and sequence mapping is performed. The continuous particle number intervals in the advancement mutation segment are extracted and the high-frequency continuous segments are marked. The synchronous advancement numbers are written into the temporary mark column. The coupling criterion for coupling path division is that the cross-coordinate overlap of the grid unit boundaries is greater than 75%. According to the coupling relationship, the advancement path set is constructed and the grid block position is combined. The multi-region grid state coupled advancement structure diagram is output.

[0031] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A comprehensive solution method for complex cross-section open channel flow velocity under a multi-point emission mechanism, characterized by: The following steps are involved: Step 1: Based on the open channel cross-sectional structural parameters, the number of disturbance sources, and the slope fluctuation coefficient, the ratio of cross-sectional coordinates to area is calculated. The difference between the distance between disturbance sources and the flow inclination is grouped, and the density of the intersection points of water level gradient and flow velocity is counted to generate a dynamic zoning structure delineation map of the watershed. Step 2: Based on the dynamic zoning structure delineation diagram of the watershed, the velocity and water level sequences of the mainstream section are interlaced and compared, the displacement and velocity angle of the boundary section are screened, and the boundary slope angle and depth jump points are sorted to generate a differential lattice driving structure set; Step 3: Based on the differential lattice drive structure set, the disturbance source coordinates and projection direction are extended, and a graph neural network is used to analyze the correlation between the disturbance path nodes and construct a mutation propagation structure diagram. The disturbance amplitude and propagation change rate are hierarchically sorted to generate a lattice synchronization instruction table for the disturbance action area. Step 4: Based on the grid synchronization instruction table of the disturbance action area, obtain the particle propagation timing and advancement difference, sort the overlapping boundary advancement time difference matrix, match and number the advancement direction density blocks, and generate a particle advancement sequence parallel calculation scheduling table; Step 5: Based on the parallel calculation scheduling table of the particle propulsion sequence, the boundary segment state transition and propulsion cycle difference are extracted, the propulsion order of the mutation zone is rearranged, the overlapping boundary synchronous propulsion section is reconstructed, and a multi-zone lattice state coupled propulsion structure diagram is generated.

2. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 1 is characterized in that: The dynamic zoning structure delineation diagram of the watershed includes zoning boundary lines, disturbance response area labels and hydraulic slope change area identification codes; the differential grid driving structure set includes grid grading identification, boundary difference area number and jump associated fragment index; the disturbance action area grid synchronization instruction table includes node matching path number, synchronous advancement level code and disturbance response grid index; the particle advancement sequence parallel calculation scheduling table includes advancement task segmentation sequence, parallel sub-block number and scheduling time step mapping table; the multi-zone grid state coupled advancement structure diagram includes grid boundary connection map, coupled advancement channel mapping and synchronization state mark field.

3. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 1 is characterized in that: The specific steps of generating the watershed dynamic partition structure delineation map are as follows: Based on the open channel cross-sectional structural parameters, the number of disturbance sources, and the slope fluctuation coefficient, the intervals between the cross-sectional coordinate point groups are measured and the area of ​​the contour closed curve is derived. The difference between the valley value of the length of the connecting line between the disturbance sources and the angle between the adjacent water flow directions is extracted and classified into intervals to generate the cross-sectional and disturbance relationship extraction results. Based on the extraction results of the relationship between the section and the disturbance, the elevation difference and gradient ratio between the water level sampling points in the longitudinal slope area are counted, the density distribution of the intersection points of the lateral velocity vectors is identified, and the overlapping areas of the turning section positions are screened to generate a dynamic zoning structure delineation map of the basin.

4. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 1 is characterized in that: The specific steps of generating the differential lattice driven structure set are: Based on the dynamic zoning structure delineation diagram of the watershed, the water level records at each section in the mainstream segment are aligned with the velocity records at the corresponding time points, and the differences between the water level and velocity at the same position in adjacent time steps are subtracted from each other. Trend comparison is calculated and the matching point sequences are labeled to generate a comparison set of mainstream segment slope changes. Based on the mainstream section slope change comparison set, the angle between the trajectory line of each control point in the boundary section and the adjacent velocity vector is extracted to obtain the control points whose angle is less than the specified angle range. At the same time, the slope curvature change value of the control point is extracted and sorted in sequence with the depth difference sequence to generate a differential lattice driving structure set.

5. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 1 is characterized in that: The specific steps of generating the grid synchronization instruction table of the disturbance action area are: Based on the differential lattice drive structure set, the disturbance source coordinate point is bound to the number of the starting line segment direction of the lattice boundary, and a fixed-step progressive line segment is added diagonally from the binding point, the displacement node coordinates are recorded, the cumulative length and end point of the line segment are calculated and marked, and a disturbance source extension path data set is generated; Based on the disturbance source extended path dataset, the local velocity data of each extended node is subtracted from the disturbance frequency change value of the adjacent node to generate a difference sequence. A graph neural network is used to extract the disturbance correlation between nodes and construct a jump propagation path graph to assist in matching the frequency transfer trajectory between nodes with a mutation threshold. The nodes in each difference sequence that are greater than the jump threshold are positionally paired, and the direction tracking mark is bound to the corresponding path identification code to generate a disturbance mutation node matching set. Based on the disturbance mutation node matching set, the disturbance amplitude peak value corresponding to each node and the propagation speed difference of the nodes in the downstream propagation path are extracted, and the amplitude values ​​in each segment are sorted, the hierarchical speed difference is processed, the flag bit reorganization and the target grid number are synchronously updated, and the grid synchronization instruction table of the disturbance action area is generated.

6. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 5 is characterized in that: The graph neural network extracts the disturbance correlation between nodes and constructs a jump propagation path graph. First, the extended nodes are regarded as independent nodes in the graph structure. A directed edge connection structure is established based on the spatial adjacency relationship and frequency change sequence between nodes. The disturbance frequency difference and spatial displacement between nodes are used as the basis for edge weight assignment. After the graph structure is generated, the disturbance correlation index between each node is iteratively calculated, and the highly correlated node pairs in the disturbance propagation chain are extracted to construct a frequency jump path index graph.

7. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 1 is characterized in that: The said adding of line segments with fixed step lengths in the diagonal direction specifically means that the starting line segment direction of the grid boundary where the disturbance source is located is used as the reference direction, the diagonal extension direction with a relative peak angle offset in the grid structure is determined, and the direction is extended outward, and the same value is added in the horizontal and vertical directions at the same time as the fixed step length in the two-dimensional coordinate system, and a line segment consistent with the initial direction is drawn at each step position, the starting point of the line segment is the end point of the previous node, and the end point is the current step coordinate point, and it is recursively deduced in sequence until it touches the boundary of the area and meets the preset distance condition. The starting and ending coordinates of the line segment need to be recorded, and the total length and end point information are updated based on the cumulative length of each step.

8. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 1 is characterized in that: The specific steps of generating the particle advancement sequence parallel calculation scheduling table are: Based on the grid synchronization instruction table of the disturbance action area, extract the grid number of each synchronization entry in the table and read the time field, extract the number by reading the instruction line and generate a timestamp and perform time labeling, calculate the particle advancement sequence difference and the node time difference, and generate a particle advancement timing difference list; Based on the particle advancement timing difference list, match the particle set corresponding to the time difference entry with the boundary unit number, read the particle ID and unit number mapping and identify the time difference aggregation and intersection area within the group, sort the row and column indexes of the time difference matrix, and generate a boundary advancement difference sorting matrix set; Based on the boundary advancement difference sorting matrix set, the frequency of particle numbers at the matrix intersections is counted, each intersection number list is counted and the frequency continuous segment is identified, the block number allocation and advancement order are adjusted, and a particle advancement sequence parallel computing scheduling table is generated.

9. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 8 is characterized in that: The particle advancement order difference refers to the order difference between the time tags corresponding to different particles on their respective advancement paths in particle advancement scheduling. Specifically, the particle advancement time tags are generated based on the time field read by the particles in the synchronization instruction table. The particles are then sorted according to the grid number and advancement path, and the difference in the advancement tag values ​​between adjacent particles is calculated.

10. The comprehensive solution method for complex cross-section open channel flow velocity under the multi-point emission mechanism according to claim 1 is characterized in that: The specific steps of generating the multi-region lattice state coupling advancement structure diagram are: Based on the particle advancement sequence parallel calculation schedule, the particle state labels and advancement cycle mark fields in each group of boundary segments are read continuously, the period sequence between the advancement start node and the end node is extracted and the difference range is compared one by one, the peak and valley value differences between the cycles are identified and the advancement offset is sorted, and a boundary segment advancement difference dataset is generated; Based on the boundary fragment propulsion difference dataset, the particle numbering sequence and time period index value combination in the propulsion offset segment are rearranged, the node numbers of the continuous mutation zone are located and the synchronous propulsion number list is extracted. The propulsion segment reorganization is combined with the synchronous lattice block path of the boundary coupling zone to generate a multi-zone lattice state coupled propulsion structure diagram.

Images