A method for optimizing equipment layout in power transmission and transformation engineering design

Abstract

This invention discloses a method for optimizing equipment layout in power transmission and transformation engineering design, specifically relating to the field of three-dimensional digital layout and construction accessibility verification of power station areas. It addresses the problem that the feasibility of the path for modular equipment from unloading to placement is difficult to determine under the superposition of constraints such as passage through doorways, clearance turning, and ground bearing capacity, and that edge-blocking and station subsidence leading to rework are common. By establishing a unified engineering coordinate benchmark and incorporating obstacle bodies, turning prohibition zones, overhead clearance constraints, and allowable base pressure maps, a placement scenario model is constructed, and transportation, lifting, turning, and placement conditions are discretized. Action sweep bodies and safety occupancy bodies are generated, connectivity search is performed to obtain candidate paths, gap spectrum tail ratio and base pressure residual ratio are calculated, and placement credibility coefficient and evidence chain are inferred. Finally, at the minimum conflict breakpoint, atomic actions are combined to generate a repair scheme, and the final layout and change record are output.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional digital layout and construction accessibility verification of power station areas, and more specifically, to a method for optimizing equipment layout in power transmission and transformation engineering design. Background Technology

[0002] In the design of new construction and expansion projects of power transmission and transformation, prefabricated or modular equipment is increasingly being used in complete units such as GIS bays, secondary combination modules, and relay protection modules. During the design phase, equipment layout and overall site planning are typically completed on a 3D / BIM model, and verification is performed based on rules such as clear channel width, enclosure boundaries, and structure avoidance. During the construction phase, a separate transportation and hoisting plan is prepared, determining unloading points, crane positions, turning radii, and temporary road conditions. Existing technologies often rely on static equipment shape envelopes for collision checks or two-dimensional planar layout for channel verification, and load-bearing capacity is often reflected through zoning experience or single-point verification, leading to an information gap between the design-side judgment of placement feasibility and the construction-side verification of placement feasibility.

[0003] However, when the module needs to enter the target foundation from the station entrance via the transportation channel and undergo attitude switching processes such as lifting, turning, and positioning, the existing static layout-based verification is difficult to cover the coupling mechanism of space occupation and site constraints throughout the entire process. Specifically, the minimum clearance between the turning section and the clearance of the gate, the sweeping boundary between the turning section and the wall and the existing structure, and the minimum clearance between the positioning section and the work space reserved around the foundation are compressed and not identified in some periods. At the same time, the equivalent base pressure formed by the deployment of the crane outriggers and the load transfer and the ground bearing zoning may exceed the limit under critical working conditions, which is not consistently expressed in the design model. If this problem is not identified in advance, the module may be unable to reach the path or be obstructed in positioning after it arrives on site, which may lead to rework such as temporary construction relocation, foundation modification or site adjustment, and bring about schedule fluctuations and on-site organization risks.

[0004] To address the aforementioned problems, a technical solution is provided. Summary of the Invention

[0005] To overcome the aforementioned deficiencies in the prior art, embodiments of the present invention provide a method for optimizing equipment layout in power transmission and transformation engineering design. This method establishes a unified engineering coordinate benchmark and incorporates obstacle bodies, slewing no-entry zones, overhead clearance constraints, and allowable base pressure maps. It then constructs a placement scenario model and discretizes transportation, lifting, slewing, and placement conditions, generating action sweep bodies and safety occupancy bodies. A connectivity search is performed to obtain candidate paths, and the gap spectrum tail ratio and base pressure residual ratio are calculated to infer the placement confidence coefficient and evidence chain. Finally, atomic actions are combined at the minimum conflict breakpoint to generate a repair scheme, and the final layout and change records are output, thus solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: S1: Read the 3D model of the module equipment, extract the shape envelope, define the entrance unloading point and the foundation landing area, convert the site object into an obstacle body, write the allowable base pressure map of the turning no-entry zone clearance constraint body, and obtain the landing scene model. S2: The placement process is discretized into a finite state sequence of transportation, lifting, rotation and placement, which limits the attitude and position. The shape envelope generates the action sweep body, and the obstacle clearance constraint body and the no-entry zone are gathered into a constraint set. S3: The candidate path set is obtained by searching the attitude connectivity from the unloading point to the target landing position. The conflict between the action sweep body and the constraint set is detected. The candidate path is calculated by the gap spectrum tail ratio and the basic pressure residual ratio to infer the landing confidence coefficient. The minimum conflict breakpoint evidence chain is output and pruned and sorted. S4: Construct an atomic action set to generate repair candidates by combining the minimum conflict breakpoint range. Apply the repair candidates to the placement scenario model and repeat step S3. Select a solution based on the placement confidence coefficient and cost index and write it into the model change record.

[0007] Furthermore, the 3D geometric model of the module equipment is read, the set of lifting points and the set of ports are extracted, the envelope of the module shape is obtained and the envelope orientation and envelope center are recorded, the constraints of transportation and hoisting methods are analyzed to generate the set of allowable postures and the set of posture switching conditions, a unified engineering coordinate benchmark is established, and the placement scene model is obtained.

[0008] Furthermore, the site enclosure boundary, existing framework, walls, roads, gates, and temporary facilities are converted into a set of obstacles and unified to a unified engineering coordinate benchmark. The entrance unloading point, target foundation placement area, transportation channel, crane position area, turning no-entry zone, and overhead clearance constraint are defined to form an allowable base pressure map and write it into the placement scenario model.

[0009] Furthermore, based on the placement scenario model, the placement process is discretized into the transportation section, lifting section, slewing section, and placement section. A set of state nodes containing working condition labels, attitude identifiers, and position identifiers is constructed, and allowed transfer boundaries are defined.

[0010] Furthermore, based on the set of state nodes, a keyframe sequence consisting of construction method action nodes and geometric event nodes is generated. The action sweep body is formed by using the module shape envelope and the lifting device shape envelope and then expanded to obtain the safety occupancy body. The obstacle body set, the passage constraint body set, the turning no-entry zone upper clearance constraint body, and the load-bearing no-entry zone form a constraint set and output the conflict determination interface.

[0011] Furthermore, a connectivity search is performed on the set of state nodes and the allowed transition boundary to generate a set of candidate paths. The safe occupancy body is called segment by segment according to the set of candidate paths, and conflict detection is performed with the constraint set. The conflict segment index and the trigger constraint object identifier are recorded to form the minimum conflict breakpoint marker.

[0012] Furthermore, the minimum signed gaps of the safe occupants and obstacles are sampled along the candidate path set to generate a gap sequence with working condition labels. Intra-phase consistency elimination and phase alignment resampling are performed to obtain a phase-aligned gap sequence. The gap spectrum tail ratio is calculated based on the low gap tail segment of the phase-aligned gap sequence.

[0013] Furthermore, the key working condition set is extracted from the candidate path set, the equivalent base pressure of the outriggers is calculated, and the allowable base pressure map is aligned to obtain the base pressure residual ratio. The gap spectrum tail ratio and the base pressure residual ratio are input into an uncertainty inferrer with monotonic constraints to obtain the landing confidence coefficient and uncertainty. Based on the landing confidence coefficient and uncertainty, the candidate path is pruned and sorted, and the evidence chain is output.

[0014] Furthermore, based on the minimum conflict breakpoint, a set of atomic actions is constructed in the local area, and the action amplitude boundary and implementation conditions are defined for each type of atomic action to form a repair candidate generation rule table.

[0015] Furthermore, a set of repair candidates is formed by combining atomic actions according to the repair candidate generation rule table. For each repair candidate, the recalculation is performed in the placement scenario model, and step S3 is reused to obtain the placement confidence coefficient and evidence chain. At the same time, the cost index is combined to complete the sorting and screening, output the list of atomic actions and evidence chain of key path nodes and write it into the model change record.

[0016] The technical effects and advantages of the equipment layout optimization method in power transmission and transformation engineering design of the present invention are as follows: This invention unifies the entire process of modular equipment in power transmission and substations from unloading to placement under the same engineering coordinate benchmark for traceable spatial simulation. It incorporates geometric obstacles, clearance requirements, no-turn access, passage boundaries, and ground bearing capacity into the same set of judgment criteria. This eliminates reliance on experience-based trial placement or repeated rework for equipment layout. Instead, it automatically forms a feasible placement path from the entrance to the foundation with the common goal of "accessible, feasible, and stable." The invention provides credibility and risk evidence at the path level, helping designers determine the feasibility of a layout scheme for construction and hoisting at the drawing stage. This significantly reduces the schedule fluctuations and organizational costs caused by temporary on-site rerouting, demolition of temporary buildings, and adjustments to station locations.

[0017] Meanwhile, in addition to feasibility assessment, this invention introduces a joint evaluation of two types of hidden risks: edge clearance and load-bearing edge contact. It reveals the low-probability, high-cost problem of "seemingly non-collision but actually very easy to get stuck or crushed" in advance. Through the chain of evidence, the risk is located to the specific conflict section, specific constraint object and key working condition. This prevents subsequent repairs from blindly expanding the scope of modification. Instead, it generates a small number of feasible adjustment actions around the minimum conflict breakpoint and quickly verifies their effectiveness. In this way, under the premise of meeting the construction method constraints, it stably outputs a final layout scheme with controllable, explainable and verifiable modification range, improving the certainty of the implementation of the design results and the efficiency of cross-disciplinary collaboration. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a method for optimizing equipment layout in power transmission and transformation engineering design according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example 1: Figure 1 This invention provides a method for optimizing equipment layout in power transmission and transformation engineering design, comprising: S1: Read the 3D model of the module equipment, extract the shape envelope, define the entrance unloading point and the foundation landing area, convert the site object into an obstacle body, write the allowable base pressure map of the turning no-entry zone clearance constraint body, and obtain the landing scene model. S2: The placement process is discretized into a finite state sequence of transportation, lifting, rotation and placement, which limits the attitude and position. The shape envelope generates the action sweep body, and the obstacle clearance constraint body and the no-entry zone are gathered into a constraint set. S3: The candidate path set is obtained by searching the attitude connectivity from the unloading point to the target landing position. The conflict between the action sweep body and the constraint set is detected. The candidate path is calculated by the gap spectrum tail ratio and the basic pressure residual ratio to infer the landing confidence coefficient. The minimum conflict breakpoint evidence chain is output and pruned and sorted. S4: Construct an atomic action set to generate repair candidates by combining the minimum conflict breakpoint range. Apply the repair candidates to the placement scenario model and repeat step S3. Select a solution based on the placement confidence coefficient and cost index and write it into the model change record.

[0021] This invention addresses the layout design of large modular equipment within power transmission and substations, treating the entire process of "equipment entering the station from the unloading point and finally landing on the foundation" as a complete accessibility problem, rather than simply checking the static collision relationships after the equipment is in place. The method first unifies the equipment's 3D model, lifting points and connection information, transportation and hoisting method requirements, and on-site objects such as perimeter walls, gate openings, roads, and temporary structures into a single engineering coordinate system. It also incorporates restrictions on slewing, overhead clearance, and ground bearing capacity into the scene model, thus forming a unified constraint expression.

[0022] Based on this, the placement process is broken down into a finite state sequence according to four working conditions: transportation, lifting, rotation, and placement. Each state is represented by an action sweep volume, which represents the continuous space occupied by the equipment in that working condition, and expands outward to form a safety occupancy volume as the construction reserve space. Then, a connectivity search is performed from the entrance to the target placement posture, and the conflict between the safety occupancy volume and the constraint set is checked segment by segment to obtain a set of candidate paths. In response to the risk of "no surface collision but easy jamming or indentation", the method further uses two types of indicators to characterize the path quality: one reflects the concentration of low clearance edge sections, and the other reflects the remaining load of the crane outriggers under key working conditions. Then, an inferrer with monotonic constraints is used to output the placement confidence coefficient and evidence chain. Finally, around the minimum conflict breakpoint located by the evidence chain, a small number of implementable atomic actions are combined to complete the local repair and verification, and the final layout plan, key nodes, action list, and traceable verification records are output.

[0023] The layout of equipment in power transmission and substation areas is often constrained by spatial conditions such as enclosure boundaries, existing structures, access roads, temporary facilities, no-entry restrictions for turning around, and overhead clearance. In addition, hoisting and transportation methods impose hard constraints on posture and position. If there is a lack of unified coordinate benchmarks and consistent scene expressions, it is difficult to transform "whether it can be entered, turned around, and landed" into calculable spatial objects and constraint relationships during the design phase. Layout judgment is likely to remain at the level of experience-based trial placement and local verification.

[0024] S101 module geometric semantic extraction and shape envelope solution.

[0025] Connectivity search and motion sweep volume generation require stable and computable geometric boundaries, as well as clearly defined slingpoint and connection port locations; otherwise, the attitude set and attitude switching conditions cannot fall into the executable space occupancy judgment.

[0026] The 3D geometric model of the modular equipment comes from the 3D component files or manufacturer model files delivered by the engineering design. The model contains component geometry and attribute information. First, the lifting point identifier and connection port identifier are read from the model attributes to form a lifting point set and a port set. The lifting point set records the spatial position and hanging direction of each lifting point, and the port set records the spatial position and port orientation of each port. When the model attributes lack lifting point identifiers, geometric supplementation identification is performed. First, structural areas with closed hole boundaries and thickened flanges around them are selected as lifting point candidates on the surface of the geometric body. Then, geometric consistency verification is performed on the candidate areas. The verification includes the closure of the hole boundary, the stability of the hole axis, and the thickness continuity around the hole. The center point of the candidate area that passes the verification is written into the lifting point set.

[0027] The shape envelope is solved using a verifiable minimization process. First, the convex hull of the module geometry is calculated, yielding a set of convex hull facets and edges. Then, a set of candidate orientations is constructed from the normal directions of the convex hull faces and the main edge directions, ensuring that no two orientations are collinear. Next, envelope evaluation is performed on each candidate orientation. The evaluation process involves rigidly rotating the module geometry under the candidate orientation, then calculating the minimum and maximum projections of the geometry along the three orthogonal directions. The projection difference forms the envelope edge length, and the product of the three envelope edge lengths forms the envelope volume. After traversing all candidate orientations, the orientation with the smallest envelope volume is selected as the envelope orientation, and the corresponding three-direction envelope edge lengths and envelope center position are recorded to form the module's shape envelope. The envelope center position is taken as the position of the arithmetic center of the rotated geometry's vertex coordinates, rotated back to the engineering coordinate reference.

[0028] Example: When the indoor station secondary combined cabin model file contains lifting lug family parameters, the lifting point set is directly extracted from the lifting lug family parameters. The convex hull surface direction mainly comes from the cabin outer shell surface and the base surface. The number of candidate orientations is limited. The orientation with the smallest envelope volume usually corresponds to the orientation where the long side of the cabin is consistent with the transportation direction. The shape envelope is used for subsequent calculation of the transportation section's occupied boundary.

[0029] S102 Construction Method Constraint Analysis Generates Allowable Attitude Set and Attitude Switching Condition Set The constraints of transportation and hoisting methods determine the family of attitudes that the module is allowed to use and the conditions under which attitude switching occurs. If the constraints of the methods are not explicitly structured, the connectivity search will generate attitude transitions that are not feasible on site, thus generating candidate paths that seem reachable but are actually impossible to place.

[0030] Read the attitude requirements and operational restrictions clauses in the transportation and hoisting procedures documents. These clauses include requirements for transport vehicle loading, spreader attachment, slewing restrictions, and placement attitude. The permissible attitude set consists of a discrete attitude family and a permissible swing range. The discrete attitude family includes at least the transportation attitude, hoisting attitude, slewing attitude, and placement attitude; the values ​​for the discrete attitude family are directly given by the procedure clauses. The permissible swing range is expressed as upper and lower angular limits around three mutually orthogonal axes. These limits are calculated from the permissible swing angles or operational restrictions in the procedure clauses, with the upper angle limit not less than the lower angle limit.

[0031] The attitude switching condition set is generated based on attitude pairs. For any attitude pair, the availability of the lifting point set under both attitudes is first verified. The verification logic is that the spatial height of the lifting point position and the reachable direction of the lifting device under both attitudes meet the construction method restrictions. Next, the orientation restrictions of the port set are verified. The verification logic is that the port orientation does not enter the prohibited orientation range during attitude switching. Then, the occupancy restrictions of the transport channel and the overhead clearance constraint are verified. The verification logic is that the occupancy outline formed by the module's outer shape envelope during attitude switching does not intrude into the channel boundary and does not intrude into the clearance constraint. After completing the above verifications, the attitude pairs and corresponding restrictions are written into the attitude switching condition set. The attitude switching condition set is directly used in step S2 to construct the allowed transitions of the finite state sequence.

[0032] S103 Establishment of Unified Engineering Coordinate Benchmark and Physicalization of Spatial Objects and Obstacles.

[0033] Design delivery models, site plans, and construction organization drawings are often based on different coordinate references. If rigid body registration is not performed and spatial objects are converted to the same coordinate system, the position of the obstacle will drift as a whole, resulting in inconsistencies between the transport channel, turning no-entry zone, and clearance constraint body and the actual site.

[0034] The coordinates of the engineering survey control network are selected as the unified engineering coordinate reference. At least three non-collinear control points are read in pairs of coordinates, consisting of source coordinates and target coordinates. The source coordinates are derived from the design model coordinates or construction organization diagram coordinates, and the target coordinates are derived from the engineering survey coordinates. Rigid body registration is performed to solve for rotation and translation. The solution process involves first calculating the geometric centers of the source control point set and the target control point set, then translating the source and target control points to their respective geometric centers. The correlation matrix between the two sets is then calculated and matrix decomposition is performed to obtain the closest orthogonal rotation. Finally, the rotation is applied to the source geometric center, and the difference between this rotation and the target geometric center is calculated to obtain the translation vector. After the solution is completed, coordinate transformation is performed on all spatial objects, including enclosure boundaries, existing structures, walls, roads, doorways, and temporary facilities.

[0035] The transformed spatial objects are uniformly converted into a set of obstacles. Walls, existing structures, and temporary facilities are converted into closed polyhedral obstacles. Roads and doorways are converted into passage constraints, which include the boundaries of passable areas and non-intrusive boundaries. Each obstacle is recorded with its planar projection outline, vertical height range, and object category identifier. When the number of control points is insufficient or the control point deviation is too large, outlier removal is performed. The removal logic is to remove outliers from largest to smallest residual and resolve until the residual meets the design coordinate accuracy requirements or there are not enough control points to continue solving.

[0036] Example: The existing station area has existing walls and gate openings. Surveyors provide the engineering coordinates of the corner points on both sides of the gate opening and the corner points of the foundation. The design model provides the model coordinates of the corresponding corner points. After rigid body registration, the wall obstacle and the gate passage constraint fall under a unified engineering coordinate reference. The subsequent transportation channel search can accurately reflect the passage conditions of the gate opening.

[0037] S104 Key Area Definition and Permissible Base Pressure Map Construction.

[0038] Connectivity search requires a clear start and end point, and it is necessary to screen the crane position and slewing restricted area under ground bearing conditions; if the permissible base pressure map or map boundary processing is lacking, the base pressure evaluation of key working conditions will result in undefined areas and interruption.

[0039] The entrance unloading point is determined by the unloading location shown in the construction organization diagram, and is recorded using spatial coordinates under a unified engineering coordinate datum. The target foundation placement area is determined by the outer contour of the foundation and the foundation elevation range, and is recorded as the volume area formed by the vertical expansion of the foundation contour.

[0040] The transport corridor is generated from the entrance unloading point to the target foundation placement area. First, the inner boundary region of the wall is extracted on the obstacle set plane projection. Then, the intersection with the road passable boundary region is obtained to obtain the basic passable area. Then, the effective clear width boundary of the doorway is trimmed, and the passable margin is reduced inward after trimming to obtain the final passable area. Subsequently, directional corridor growth is performed with the entrance unloading point as the starting point and the center line of the target foundation placement area as the guiding curve to output the transport corridor.

[0041] The crane positioning zone is obtained by intersecting the outer radius ring of the target foundation placement area with the accessible area. First, based on the upper and lower limits of the working radius in the lifting method constraints, a ring-shaped outer expansion is performed on the target foundation placement area to generate candidate positioning rings. Then, the areas covered by the slewing restricted area and the load-bearing restricted area are removed from the candidate positioning rings, and the remaining area is defined as the crane positioning zone. The slewing restricted area is generated by the geometric expansion of the enclosure boundary, the existing energized area, and the owner's restricted area. The overhead clearance constraint is obtained by combining the vertical capping body formed by the height range of the obstacle and the motion envelope formed by the boom's moving envelope.

[0042] The allowable base pressure map is generated from geological survey bearing capacity data or paving scheme bearing capacity data. The bearing capacity values ​​and locations of sampling points within the station area are written into the sampling point set, with the bearing capacity values ​​in pressure units. Triangulation is performed on the sampling points to form a triangular mesh. Inside any triangle, the allowable base pressure is taken as a linear interpolation of the allowable base pressure at the three vertices proportional to their area. Outside the triangular mesh coverage area, the allowable base pressure is taken as the allowable base pressure of the nearest sampling point or as a constant value of the bearing zone. Boundary backoff rules are fixedly recorded in the allowable base pressure map. After completion, the module shape envelope, lifting point set, port set, allowable attitude set, attitude switching condition set, obstacle set, entrance unloading point, target foundation placement area, transport channel zone, crane station zone, turning no-entry zone, overhead clearance constraint body, and allowable base pressure map are written into the placement scene model.

[0043] Step S1 unifies the geometry and connection point semantics of the module equipment, the constraints of transportation and hoisting methods, the spatial objects of the site and the access nodes under the same engineering coordinate reference, forming a landing scene model that includes obstacles, entrance unloading points, target foundation landing areas, transportation channels, crane positions, slewing no-entry zones, overhead clearance constraints and allowable base pressure maps, so that subsequent judgment objects have a unified spatial reference and traceable constraint sources.

[0044] The placement process is not a static placement of a single pose point, but an occupation trajectory formed by the continuous evolution of working conditions such as transportation, lifting, rotation, and placement. If only the final state geometry is checked for collision, there may still be edge-blocking or unreachable positions in areas such as doorway corners, rotation boundaries, and placement approaches. Therefore, it is necessary to explicitly express the working condition semantics, attitude switching, and continuous occupation space as a determinable state structure.

[0045] S201 Working Condition Semantic Discretization and Working Condition Attitude Subset Generation.

[0046] The placement of prefabricated modules includes four types of working conditions: ground transportation, lifting off the ground, aerial rotation, and precise positioning. The attitude constraints for these four types of working conditions come from different sources. When the boundaries of the working conditions are unclear, it is easy to use the transportation attitude for the rotation section or the rotation attitude for the placement section, which will lead to the distortion of subsequent conflict detection results.

[0047] The placement scenario model provides a set of allowed postures and a set of posture switching conditions. The transportation and hoisting method constraints provide a mapping relationship between the working condition clauses. First, based on the working condition clauses, the set of allowed postures is divided into a subset of transportation postures, a subset of hoisting postures, a subset of slewing postures, and a subset of placement postures. The division rule is that when the posture item identifier matches the working condition clause label, it enters the corresponding subset. Then, based on the set of posture switching conditions, the set of posture switching condition subsets between adjacent working conditions is extracted. The posture switching condition subsets correspond to the three directions of transportation to hoisting, hoisting to slewing, and slewing to placement, respectively. The extraction rule is that the posture before switching belongs to the posture subset of the previous working condition and the posture after switching belongs to the posture subset of the next working condition, while also satisfying the lifting point availability and prohibited orientation clauses in the working method constraints.

[0048] S202 State Node Construction and Allowed Transition Boundary Definition.

[0049] Connectivity search requires advancing candidate paths on state nodes. State nodes must simultaneously carry a condition label, an attitude availability domain, and a location reachable domain. Allowable transition boundaries must simultaneously satisfy attitude switching conditions and location domain connectivity conditions; otherwise, candidate paths may have breakpoints where they are geometrically connected but attitude cannot be switched.

[0050] Allowable transition boundaries are established one by one according to state pairs. First, the attitude switching conditions are verified. If the attitude switching conditions are met, the position domain verification is entered. The position domain verification uses the union node sequence formed by the transport vehicle outline change nodes and the channel corner nodes as the basis for segment division. The reachable domain of the state position is read segment by segment, and the continuity of the first and last ends is determined. If the continuity is established, the allowed transition edge is written. The path cost is the sum of the working condition advancement step cost and the planar advancement distance cost on the same scale. The heuristic cost is the sum of the planar projection distance from the current state node to the terminating state node and the remaining working condition steps. The conversion coefficient is calibrated offline and satisfies that the heuristic cost does not exceed the actual remaining cost.

[0051] The transfer boundary is established under two conditions: the attitude condition is any switching pair in the subset of adjacent working condition attitude switching conditions, and the position condition is the existence of a connecting corridor between two types of position markers. The connecting corridor is determined by geometric intersection and superposition of the travel outline. The connecting occupancy zone is constructed by the outer outline of the transport vehicle's turning outline and the outer outline of the module's shape envelope in the transport attitude. The connection from transport to lifting position is determined to be established when the connecting occupancy zone and the crane position zone have a non-empty overlap. The connecting zone from lifting to turning position is constructed by the spatial projection of the hook vertical line and the module lifting point set. The connection from lifting to turning position is determined to be established when the position connecting zone and the turning prohibition zone in the turning attitude do not overlap. The connection from turning to landing position is determined to be established by the overlap relationship between the module's shape envelope projection at the turning end point and the target foundation landing area projection.

[0052] S203 State constraint domain generation and lifting point lifting tool consistency verification.

[0053] The sweeping motion needs to cover the combined occupancy of the module's outer envelope and the spreader's outer envelope. When the set of lifting points and the spreader's attachment method are inconsistent, the sweeping motion will miss the spreader's occupancy or introduce an unattachable posture on site, resulting in missed or false detections in conflict detection.

[0054] For each state node, a state constraint domain is generated. The state constraint domain consists of an attitude availability domain and a position reachability domain, with an additional lifting point availability constraint. The lifting point availability constraint calculates the included angle based on the lifting point set and the hook main direction clause in the lifting method constraint. The angle calculation order is as follows: first, take the lifting point direction vector and the hook main direction vector, then calculate the dot product of the two vectors, then calculate the product of the magnitudes of the two vectors, then divide the dot product by the magnitude product to obtain the cosine value, and then take the inverse cosine of the cosine value to obtain the included angle. When the included angle falls within the allowable range given by the method constraint, the lifting point is included in the set of available lifting points for the working condition. The lifting tool shape envelope comes from the parameterized lifting tool template corresponding to the lifting tool model and sling arrangement recorded in the placement scene model. The lifting tool template outputs the sling unfolded contour and the hook envelope contour. The set of available lifting points for the working condition and the lifting tool template jointly determine the position and orientation of the lifting tool shape envelope under a unified engineering coordinate reference.

[0055] Example: In the four-point hoisting of the prefabricated cabin, four slings are used to attach four lifting lugs in the lifting section. The availability of the lifting points is checked and lifting lugs that are obstructed and cause the incident angle of the slings to not meet the construction method range are removed. The outer envelope of the lifting tool is updated as the slings unfold. The slewing section and the placement section use the same outer envelope of the lifting tool and keep the attachment points consistent.

[0056] S204 Action Sweep and Safe Occupation Generation.

[0057] Connectivity search requires conflict detection of the continuous occupied space for each working condition. Single-point trajectories cannot express the sweep boundaries of the module shape envelope and the spreader shape envelope during the action. The action sweep volume must be constructed by a verifiable discrete keyframe sequence to ensure the feasibility of the project.

[0058] A keyframe sequence is generated for each state node. The keyframe sequence is derived from the union of construction method action nodes and geometric event nodes. Construction method action nodes include unloading and positioning, lifting off the ground, slewing start, slewing end, approaching positioning, and positioning. Geometric event nodes include transportation channel corners, passage through doorways, proximity to the perimeter wall, proximity to the framework, and proximity to the foundation boundary. Intermediate frames are generated between adjacent keyframes using attitude linear interpolation and position linear interpolation. The interpolation constraints satisfy that the attitude always belongs to the subset of the corresponding working condition attitude and the position always belongs to the reachable domain of the corresponding position.

[0059] The motion sweep body is constructed from the frame-by-frame occupancy union of the keyframe sequence. The frame-by-frame occupancy is obtained by merging the occupancy body after the spatial transformation of the module shape envelope in the current frame and the occupancy body after the spatial transformation of the spreader shape envelope in the current frame. The safety occupancy body is obtained by expanding the shape of the motion sweep body according to the work retention distance in the construction method constraint. The shape expansion is offset outward along the normal direction of the sweep body surface and fills the cavity formed by the outward offset. The outward offset distance is the work retention distance.

[0060] S205 constraint set fixing and conflict determination interface output.

[0061] When advancing the candidate path segment by segment in step S3, a unified conflict determination interface is required. The conflict determination interface must fix the determination object and determination rules to avoid the misuse of the turning no-entry zone rule in the transportation segment or the omission of the upper airspace constraint body rule in the turning segment.

[0062] The constraint set is directly read and fixed from the landing scene model. The constraint set includes the obstacle set, the passage constraint set, the turning prohibition zone, and the overhead clearance constraint. The load-bearing prohibition zone is obtained by combining the area outside the boundary of the allowable base pressure map and the prohibited area marked in the data. The output conflict determination interface includes three types of inputs and three types of determination relationships. The inputs include the state node set and the allowable transfer boundary, the set of safe occupants in each state, the constraint set and the load-bearing prohibition zone. The determination relationships include the intersection determination of safe occupants and obstacles, the intersection determination of safe occupants and overhead clearance constraint, and the determination of safe occupants exceeding the boundary of the passage constraint under transportation conditions. At the same time, the overlap relationship between the load-bearing prohibition zone and the crane station zone is used as the determination of the station validity.

[0063] Step S2 discretizes the placement process according to the working condition semantics and constructs a finite state sequence. For each state, a constraint domain is generated for the module attitude parameters and position parameters. Based on the module envelope and the shape of the lifting device, an action sweep body and its safety occupancy body are generated. At the same time, the obstacle body, the clearance constraint body, the turning no-entry zone and the load-bearing no-entry zone are solidified into a constraint set, so that the conflict relationship between the continuous occupancy space and the constraint object becomes a unified judgment criterion.

[0064] After the state structure and continuous occupied space are clear, two key judgments still need to be answered: one is whether the path is connected and does not trigger hard conflicts throughout the entire process, and the other is whether there is an unstable risk of seemingly no collisions but there is edge clearance and load-bearing edge. If there is a lack of comparable risk characterization and traceable location, it is difficult to sort and select candidate paths, and it is also difficult to accurately locate the problem to the specific segment and triggering constraint.

[0065] Step S3 selects the gap spectrum tail ratio and the base pressure margin ratio as inputs for comprehensive analysis because the failure causes during the placement process are mainly concentrated in two independent but often simultaneous risk links in engineering: one comes from the low clearance edge-grabbing section between the continuous occupied space and the obstacle boundary, which appears to be non-collision but causes jamming or rubbing at doorway corners, frame edges, turning boundaries, etc. The gap spectrum tail ratio can stably characterize the concentration and intensity of the low clearance tail segment, avoiding being dragged by the accidental error of the minimum clearance at a single point; the other comes from the edge-grabbing condition between the crane station and the ground bearing capacity, which appears to be that the station is nominally achievable but the outriggers collapse or the bearing capacity is insufficient under critical conditions. The base pressure margin ratio expresses the bearing edge-grabbing risk with the remaining degree of the most unfavorable critical condition, avoiding the use of static zoning or single station judgment to cover up the most unfavorable condition. Both cover "spatial obstruction risk" and "bearing capacity instability risk" respectively. They are dimensionless and directly correspond to the constraints of the construction method. They are highly complementary. When making comprehensive inferences, they can compare risks from different sources, such as low clearance and low bearing capacity, on the same credible scale. They can also clearly distinguish the source of risk in the chain of evidence, thereby improving the discrimination stability of the pruning order and reducing the scope of blind changes during repair.

[0066] S301 Candidate Path Connectivity Search and Segment-by-Segment Conflict Detection.

[0067] The landing scenario model has provided the entrance unloading point, the target base landing area, the constraint set and the allowable base pressure map. Step S2 has provided the state node set, the allowable transfer boundary, the safe occupancy body set and the conflict determination interface. The connectivity search needs to advance the starting state node corresponding to the entrance unloading point to the ending state node corresponding to the target base landing area, while ensuring that each safe occupancy body does not trigger a constraint set conflict.

[0068] A shortest path search with a heuristic function is used to generate a candidate path set on the state graph composed of the set of state nodes and the allowed transition boundary. The cumulative path cost of the current state node is defined as the normalized value of the number of allowed transitions that have occurred multiplied by a first superposition coefficient, plus the normalized value of the advanced planar distance multiplied by a second superposition coefficient. The normalization benchmarks are the minimum feasible number of transitions from the starting point to the ending point and the reference channel length, respectively. The heuristic function is defined as the normalized value of the minimum remaining number of transitions from the current state node to the terminating state node multiplied by a first superposition coefficient, plus the normalized value of the straight-line distance between the two nodes in the planar projection multiplied by a second superposition coefficient. The minimum remaining number of transitions is taken as the lower bound of the state layer, and the straight-line distance in the planar projection is taken as the geometric shortest lower bound. The superposition coefficient is obtained by offline calibration, and the heuristic function is constructed with the lower bound to keep it from exceeding the actual remaining cost. During the search process, each allowed transition is regarded as a segment of advancement. When advancing a segment, the corresponding safe occupancy volume is read and conflict detection is performed.

[0069] Conflict detection employs a two-stage decision-making process. The first stage uses a hierarchical bounding volume tree for coarse screening, eliminating non-intersecting objects based on the intersection relationship between the bounding box of the safety occupant and the bounding box of the constraint set object. The second stage performs fine-tuning on the coarsely screened objects. For separation cases, the minimum separation distance is obtained by solving the closest convexity distance problem, and for intersection cases, the penetration depth is obtained by solving the penetration depth problem and recorded as a negative clearance. The transportation section additionally performs boundary crossing judgment on the set of passage constraint bodies. Boundary crossing judgment is triggered when the outer contour of the plane projection of the safety occupant does not completely fall into the passage boundary.

[0070] Example: In the scenario of passing through a doorway, the state diagram generates two branch paths in the transportation section. One path pushes the outer contour of the safety occupant body into the boundary of the doorway, triggering the boundary violation judgment of the passage constraint body. During the search process, the boundary violation segment is marked as a conflict segment and the expansion of the corresponding branch is stopped. The other path passes near the center line of the doorway and enters the crane station area. During the search process, the corresponding state node sequence is written into the candidate path set.

[0071] The candidate path set and segment-by-segment conflict detection markers provide the sampling segment index for subsequent gap sequence calculations, and provide the conflict segment location anchor points required for the evidence chain.

[0072] S302 Signed gap sampling and stage-aligned gap sequence generation.

[0073] Project placement failures are often triggered by local edge-grabbing intervals. These edge-grabbing intervals are short in duration and concentrated in a small number of working conditions. Simply using whether there is a conflict cannot express the edge-grabbing trend and stage attribution. The clearance sequence needs to record the minimum signed clearance of each segment with stage labels and then reshape and align it.

[0074] For each candidate path in the candidate path set, each safe occupancy body is traversed in the order of the state node sequence, and the minimum signed clearance is obtained by taking the set of rigid obstacles in the constraint set as the sampling object. The order of obtaining the minimum signed clearance is as follows: first, calculate the separation distance of each obstacle after coarse screening and take the minimum separation distance; then, calculate the penetration depth of each intersecting obstacle and take the negative number of the penetration depth as the clearance value; finally, take the minimum value among all obstacle clearance values ​​as the segment clearance. The segment clearance and the segment working condition label are written together into the clearance sequence with stage label.

[0075] Anomaly removal employs intra-stage consistency rules, which are calibrated offline and then permanently recorded. Within the same working condition label, the median trend of adjacent gap sequences is first calculated, followed by the deviation of each gap from the median trend. When the deviation exceeds a preset deviation limit, the corresponding gap is replaced with the median trend value. Stage alignment uses intra-segment arc length parameter normalized resampling. First, a first resampling density sequence is generated based on the keyframe sequence spacing, and then a second resampling density sequence is generated based on the channel rotation rate. Then, the two densities are compared point-by-point at the same location, and the higher density is selected to form the execution density sequence. The stage-aligned gap sequence is output according to the execution density sequence, while maintaining the original working condition label for each gap element.

[0076] The dimension of the clearance element is length. A positive clearance element indicates that the safe occupant is separated from the boundary of the obstacle, while a negative clearance element indicates that the safe occupant intrudes into the boundary of the obstacle.

[0077] S303 gap spectrum tail ratio calculation.

[0078] The risk of edge-grabbing needs to reflect the concentration and intensity of low-gap segments. The minimum clearance at a single point is easily affected by local discrete errors. The gap spectrum tail ratio uses the low-gap tail segments as statistical objects, which can stably characterize the edge-grabbing interval.

[0079] The clearance datum is taken from the operational retention distance clause in the transportation and hoisting method constraints, and the upper bound of the modeling alignment error under the unified engineering coordinate datum is superimposed to form the clearance datum. The dimension of the clearance datum is length and takes a positive value. The non-negative clearance set is extracted from the stage alignment clearance sequence and sorted by value from smallest to largest. After sorting, the value corresponding to the preset quantile is taken as the tail boundary value. The preset quantile is obtained by offline calibration and written into the construction method rule configuration. The clearance element index not greater than the tail boundary value is defined as the tail set.

[0080] The tail average clearance is the arithmetic mean of the clearance elements in the tail set; the gap spectrum tail ratio is the ratio of the tail average clearance to the net clearance baseline and is subject to interval constraints. The interval constraints limit the gap spectrum tail ratio to a preset stable interval to ensure that the inferr input does not diverge; when the non-negative clearance set of the stage-aligned clearance sequence is empty, the tail average clearance is taken as the deepest penetration value in the clearance sequence, and the gap spectrum tail ratio is subsequently taken as a negative value while maintaining the dimensionless property after dimensional cancellation.

[0081] The gap-to-tail ratio is dimensionless. A negative gap-to-tail ratio indicates that the tail segment has a tendency to invade or is close to invading, while a positive gap-to-tail ratio indicates that the tail segment remains separated.

[0082] Example: In the scenario of the fence edge, the candidate paths do not intersect explicitly, but the stage alignment gap sequence has a continuous low gap segment in the transportation segment. The tail set covers the segment, and the average gap of the tail is lower than the clearance reference. The gap spectrum tail ratio decreases accordingly and the candidate path is moved from the front to the back of the priority queue.

[0083] The gap spectrum tail ratio improves the risk of edge-grabbing segments from a single-point extreme value to the intensity of tail segments, which can form a complementary input with the load-bearing index.

[0084] S304 base pressure ratio calculation.

[0085] Meeting the clearance and no-entry requirements for the crane position does not mean that the load-bearing capacity meets the requirements. Under critical operating conditions, the outrigger reaction force and grounding area determine the equivalent base pressure. The equivalent base pressure needs to be aligned with the allowable base pressure map in spatial position to form a redundancy expression.

[0086] Extract the key working condition set from the key frame sequence of the candidate path. The key working condition set includes the key frames corresponding to lifting off the ground, slewing start, slewing end, landing approach, and landing in place. For each key working condition, read the crane station point, outrigger grounding point coordinate set, and outrigger pad grounding area set. The outrigger grounding point coordinate is in the dimension of length, and the outrigger pad grounding area is in the dimension of area and takes a positive value.

[0087] The outrigger reaction force is solved using non-negative constraint least squares. The equation system consists of vertical force balance and moment balance in two orthogonal directions. The known quantities are the module weight, the lifting gear weight, and the lever arm geometry under key working conditions. The unknown quantities are the reaction forces of each outrigger, and the reaction forces are constrained to not be negative. When the equation system is not suitable, the solution process introduces a smoothing constraint on the change of outrigger reaction force and maintains the non-negative constraint. The smoothing constraint strength is obtained by offline calibration and written into the construction method rule configuration.

[0088] The calculation order of equivalent base pressure is as follows: first, obtain the outrigger reaction force, then divide the outrigger reaction force by the grounding area of ​​the corresponding outrigger pad to obtain the equivalent base pressure. The dimension of equivalent base pressure is pressure. The allowable base pressure is obtained by querying the plane coordinates of the outrigger grounding point on the allowable base pressure map. The dimension of allowable base pressure is pressure. The ratio of equivalent base pressure to allowable base pressure is calculated for each critical working condition. The most unfavorable ratio for the candidate path is the maximum value of the ratios of all critical working conditions and all outriggers. The base pressure residual ratio is one minus the most unfavorable ratio. The base pressure residual ratio is dimensionless. A negative base pressure residual ratio indicates a tendency to exceed the limit when touching the edge.

[0089] When the base pressure map is undefined at the outrigger grounding point, the corresponding outrigger position is written into the load-bearing restricted area marker set, and the base pressure residual ratio is directly set to a negative range. At the same time, the uncertainty marker is written into the evidence chain.

[0090] The base pressure residual ratio elevates the risk of bearing edge contact from static partition judgment to the expression of the most unfavorable residual degree under critical working conditions, and can jointly cover the two failure modes of edge contact and edge contact with the gap spectrum tail ratio.

[0091] S305 Layout Credibility Coefficient Inference, Pruning and Ordering, and Evidence Chain Output.

[0092] The gap tail ratio reflects the risk of edge-grabbing segments, while the base pressure residual ratio reflects the risk of bearing edge-grabbing. The two are complementary. Simple linear synthesis cannot express the uncertainty under high-conflict conditions. The inference process needs to output the placement confidence coefficient and uncertainty at the same time and have monotonic constraints.

[0093] An uncertainty inferrer with monotonic constraints is constructed. The inferrer takes the gap spectrum tail ratio and the basis pressure residual ratio as inputs. The inferrer adopts an evidence fusion structure and outputs three types of quality assignments, corresponding to feasible proposition quality, infeasible proposition quality, and uncertain proposition quality, respectively. The quality assignment mapping function adopts a piecewise monotonic function. The piecewise inflection point and slope are obtained by offline calibration and written into the rule configuration. The mapping function satisfies that the quality of feasible propositions does not decrease and the quality of infeasible propositions does not increase when the input increases.

[0094] Evidence fusion adopts the Demst combination rule. When the combination conflict quality exceeds the preset conflict limit, the conflict quality is transferred to the uncertain proposition quality and the sum of the three types of quality is kept to one, thereby avoiding numerical instability caused by the combination denominator approaching zero. The placement confidence coefficient is taken as the sum of the feasible proposition quality and the uncertain proposition quality multiplied by the conservative allocation coefficient. The uncertainty is taken as the uncertain proposition quality. The conservative allocation coefficient is fixedly recorded by offline calibration and its value ranges from zero to one. Both the placement confidence coefficient and the uncertainty are non-negative and do not exceed one.

[0095] Pruning and priority ranking are determined by the lower bound of the placement confidence coefficient and the upper bound of the uncertainty. The determination threshold is derived from offline calibration and is consistent with the risk calibrator of the construction method. Candidate paths entering the priority queue are arranged in descending order of placement confidence coefficient, and in the case of the same placement confidence coefficient, they are arranged in ascending order of uncertainty.

[0096] The minimum conflict breakpoint and evidence chain output follow a unified positioning rule. When a geometric conflict exists, the segment index that triggers the conflict detection earliest is taken as the minimum conflict breakpoint. When there is no geometric conflict but there is a risk of grazing or touching the edge, the minimum clearance segment index and the index of the most unfavorable critical working condition of the base pressure are jointly recorded as the minimum conflict breakpoint. The evidence chain includes at least the candidate path identifier, the minimum conflict breakpoint segment index, the trigger constraint object identifier, the trigger constraint type, the stage alignment clearance sequence segment index, the critical working condition index, the support leg index, the load-bearing forbidden zone marker, and the uncertainty marker.

[0097] Example: When the crane station crosses a weak section of the paving, the candidate path does not geometrically trigger the intersection of obstacles, but the base pressure ratio is negative and the base pressure map alignment result shows that the outrigger falls into the load-bearing restricted area. The inferrer increases the quality of the infeasible proposition and outputs a higher uncertainty. After the pruning rule is triggered, the candidate path is moved into the pruning set. The evidence chain records the outrigger index and the most unfavorable critical working condition index for step S4 to generate the station translation atomic action.

[0098] The placement confidence coefficient and uncertainty transform the candidate path set into a sortable, pruning, and traceable decision output, and anchor the minimum conflict breakpoint to a specific segment and specific constraint evidence, which facilitates step S4 in generating repair candidates within a local range.

[0099] Step S3 uses the entrance unloading point and the target landing posture as endpoints to perform connectivity search and performs conflict detection on the action sweep body to form a candidate path set. It samples along the candidate paths to obtain the gap sequence with stage labels and calculates the gap spectrum tail ratio. At the same time, it calculates the equivalent base pressure under the key working conditions and aligns the allowable base pressure map to obtain the base pressure ratio. Then, it infers the landing confidence coefficient and uncertainty and outputs the minimum conflict breakpoint and evidence chain to realize the pruning and ranking of candidate paths and risk attribution.

[0100] When the chain of evidence points to the minimum conflict breakpoint and the triggering constraint object, the key to layout repair is not to expand the scope of modification, but to propose a small number of on-site adjustments based on the conflict mechanism, and to ensure that the adjustments do not introduce new issues such as passage overstepping, turning restrictions, or load-bearing problems. Therefore, it is necessary to structure the adjustment actions into a set of actions that can be combined, screened, and verified, so that the repair process has a controllable engineering feasibility boundary.

[0101] S401 Minimum Conflict Breakpoint Local Range Limitation and Construction of Adjustable Object Set.

[0102] The evidence chain output in step S3 points to the conflict segment index and the trigger constraint object identifier. If the local range is too large, it will introduce irrelevant objects and increase the number of repair candidates. If the local range is too small, it will miss adjustable objects and cause the repair to fail.

[0103] The process involves reading the minimum conflict breakpoint marker in the evidence chain, locating the corresponding conflict segment index associated with the safe occupancy body, and locating the trigger constraint objects recorded in the evidence chain to form a set of trigger constraint objects. A local range volume is generated centered on the outer boundary of the safe occupancy body. This local range volume is generated by first taking the discrete boundary point set of the outer surface of the safe occupancy body, then performing neighborhood expansion on the boundary point set with an outer radius, and merging the expansion results into voxels. The outer radius is determined according to fixed calculation rules: first, the operation retention distance is extracted from the constraints of the transportation and hoisting methods; then, the upper bound of the modeling alignment error is extracted from the control point registration residual sequence during the establishment of the unified engineering coordinate reference; finally, the two are added together to obtain the outer radius. The outer radius has the dimension of length and is positive. The operation retention distance comes from the constraints of the transportation and hoisting methods, and the upper bound of the modeling alignment error comes from the upper bound of the control point registration residual during the establishment of the unified engineering coordinate reference. The upper bound of the modeling alignment error has the dimension of length and is positive. The spatial overlap between the local range volume and the transportation channel with the crane station and the target foundation placement area is taken as the effective domain, which is used to limit the range of action. Based on the effective domain, an adjustable object set is constructed. The adjustable object set includes the positioning reference point module, the crane station point, the adjacent facility object identification channel boundary node, and the inclusion rule is that the object boundary and the effective domain have spatial overlap or the shortest distance from the object boundary to the boundary of the trigger constraint object is less than the outer radius. The unit of the shortest distance is length.

[0104] When the outer radius is a positive length, the effective domain is a closed domain, the set of adjustable objects is a finite set and the member identifiers come from the placement scene model.

[0105] S402 atomic action set construction and action rule table solidification.

[0106] The risks revealed by the chain of evidence may come from three modes: passing over obstacles, grazing, bearing, and touching the edge. Atomic actions need to cover all three modes and maintain on-site feasibility. The lack of action range boundaries and availability conditions will prevent the remediation candidates from being implemented.

[0107] Within the effective domain, a set of atomic actions is constructed, including basic micro-movement actions, module orientation adjustment actions, crane positioning translation actions, and temporary passage clearance actions. An action parameter set is generated for each type of atomic action, containing planar displacement and angular deflection. The planar displacement is measured in length and its value range is limited to the corresponding action amplitude boundary. The angular deflection is measured in angle and its value range is limited to the allowable attitude set coverage area. The action amplitude boundary of the basic micro-movement action is derived from the civil engineering allowable offset caliber and the target foundation placement area boundary. The civil engineering allowable offset caliber comes from the foundation construction drawings or secondary design change clauses. After the basic micro-movement, the placement reference point remains within the target foundation placement area, and the connection direction of the port set still satisfies the reserved passage direction constraint. The action amplitude boundary of the module orientation adjustment action is derived from the allowable attitude set and attitude switching condition set. After the module orientation adjustment, the attitude still belongs to the allowable attitude set, and the attitude switching condition is still satisfied when switching between adjacent working conditions. The lifting point set still satisfies the hook main direction angle constraint after adjustment, and the angle constraint comes from the transportation and lifting method constraint clauses. The boundary of the crane's lateral movement is determined by the combined constraints of the crane's position, the slewing restriction zone, and the load-bearing restriction zone. After the crane's position point is moved, it must still fall within the crane's position zone but not the slewing restriction zone, and the outrigger grounding point set must not overlap with the load-bearing restriction zone. The lifting radius is determined using a verifiable calculation sequence: first, obtain the plane coordinates of the crane's slewing center; then, calculate the plane projection coordinates of the centroid of the lifting point set; finally, calculate the plane distance between the two as the lifting radius. The allowable range of the lifting radius is derived from the lifting method constraints. The boundary of the temporary construction passage clearance action is determined by the allowable relocation zone of the temporary construction and the road portal passage boundary. The allowable relocation zone is derived from the temporary construction layout plan or on-site management regulations. After relocation, the outer envelope of the temporary construction must completely fall within the allowable relocation zone without encroaching on the passage constraint boundary or the projection area of ​​the overhead clearance constraint. The boundary of the range and availability conditions for each type of action are written into the action rule table, which is used for subsequent combination generation and filtering.

[0108] Example: When the passage through the doorway is triggered by exceeding the boundary, the temporary construction passage clearance action first moves the temporary construction fence along the outside of the road to the boundary of the allowed relocation area. After the passage boundary is restored, the module orientation adjustment action is executed to align the long side of the module with the center line of the doorway. The hoisting point set maintains the same attachment point, the crane station position remains unchanged, and the on-site actions are consistent with the evidence chain triggering object.

[0109] S403 Action Trigger Mapping Driven Combination Generation and Quick Filtering.

[0110] A single atomic action cannot simultaneously eliminate the risk of grazing and the risk of bearing the edge. Action combinations need to be generated in a directional manner according to the trigger constraint type to control the search space, and combinations that are necessarily infeasible should be screened out before recalculation to reduce computational costs.

[0111] Action trigger mapping is established based on the trigger constraint types in the evidence chain. When the trigger constraint type is "crossing the boundary," the temporary passage yielding action and module orientation adjustment action are prioritized. When the trigger constraint type is "grabbing an obstacle," the module orientation adjustment action and basic micro-movement action are prioritized. When the trigger constraint type is "carrying a restricted area marker," the crane positioning translation action is prioritized. The action trigger mapping only selects actions that meet the availability conditions from the action rule table to enter the enumeration set. Action combination generation adopts hierarchical enumeration and sequential solidification rules. The sequential solidification rule stipulates that actions related to the passage boundary are executed first, followed by actions related to the positioning, then actions related to the landing reference point, and finally actions related to orientation. Sequential solidification is used to avoid actions canceling each other out and reduce repeated combinations.

[0112] The rapid screening process includes two types of necessary condition determinations. The first type is the determination of local minimum clearance improvement. This is achieved by applying the action combination to the discrete point set of the safe occupant boundary within the effective domain and calculating the minimum nearest distance from the discrete point set to the boundary of the triggering constraint object. A positive minimum nearest distance indicates local separation, while a negative minimum nearest distance indicates local intrusion. The second type is the determination of local load-bearing feasibility. This is achieved by applying the action combination to the outrigger grounding point set and querying the allowable base pressure map to obtain the allowable base pressure set. The most unfavorable equivalent base pressure of the critical working condition before recalculation in step S3 is then used as a conservative upper bound for comparison. The source of the upper bound is fixed as the most unfavorable value in the equivalent base pressure sequence output in step S3. Action combinations that satisfy both types of necessary conditions are written into the repair candidate set. The repair candidate set records the action combination identifier, action parameter set, trigger evidence fragment index, and screening determination results.

[0113] Example: When the perimeter of the wall is brushed and accompanied by a no-entry zone mark, the action trigger mapping incorporates the module orientation adjustment action and the crane position translation action into the same combination. The combination of the position translation and the outrigger falling into the boundary of the slewing no-entry zone is quickly filtered out first, and then the combination of the position translation and the outrigger falling into the undefined area of ​​the allowable base pressure map is also filtered out. The remaining combination enters the repair candidate set.

[0114] S404 Repair and Placement Scene Model Generation and Step S3 Recalculation and Final Solution Selection.

[0115] Rapid screening only verifies local necessary conditions. Complete feasibility needs to be reviewed at the full-path scale. The review criteria need to be consistent with step S3 to ensure that the sorting and evidence chain are traceable. At the same time, the selection of the scheme needs to control the scope of modification to avoid introducing disputes over weight superposition.

[0116] For each repair candidate in the repair candidate set, the action parameter set is applied to the corresponding object in the placement scene model to obtain the repair placement scene model. The application method follows the magnitude boundary and availability conditions of the action rule table and maintains a unified engineering coordinate reference. Step S3 is repeated on the repair placement scene model to obtain the placement confidence coefficient uncertainty and evidence chain of the repair candidate path set. A cost index vector is calculated for each repair candidate. The cost index vector consists of the action quantity cost, maximum displacement cost, and affected area cost. The action quantity cost is the number of atomic actions in the action combination, the maximum displacement cost is the maximum value of all planar displacements in the action parameter set, and the affected area cost is the area of ​​the union of the effective domain and the boundary of the action object projected onto the horizontal plane.

[0117] The final solution employs a two-tiered criterion. The first criterion prioritizes repair candidates with higher placement confidence coefficients and lower uncertainty. The second criterion, when the first criterion is inconclusive, selects the candidate with the smaller cost index vector in lexicographical order. The lexicographical order comparison is fixed as follows: cost of number of actions takes precedence over cost of maximum displacement, which in turn takes precedence over cost of affected area. Simultaneously, an evidence chain consistency constraint is added, requiring that the minimum conflict breakpoint marker in the evidence chain improves after repair compared to before repair. The improvement criterion is fixed as the conflict segment index shifting, the trigger constraint type changing from intersection to separation, or the disappearance of the load-bearing forbidden zone marker. The placement confidence coefficient and uncertainty range from zero to one, and all components of the cost index vector are non-negative.

[0118] S405 solution output and model change record writing.

[0119] On-site implementation requires converting the plan into an executable list that supports verification and traceability. The output content must be bound one-to-one with the module object identifier; otherwise, the review process cannot locate the source of the action and the evidence of risk.

[0120] The system outputs the final path key node sequence, including key nodes for the transportation section, crane station location, slewing start and end key nodes, and placement key nodes. It also outputs a list of atomic actions, containing atomic action type identifiers, action parameter sets, amplitude and boundary satisfaction markers, and availability condition satisfaction markers. Finally, it outputs a chain of evidence, including minimum conflict breakpoint markers, trigger constraint object identifiers, trigger constraint type, stage alignment, clearance sequence, fragment index, key working condition index, outrigger index, and load-bearing restricted area marker changes. The path key node sequence, atomic action list, and evidence chain are bound to module object identifiers and written into the model change record. This record includes the model identifier for the placement scenario before repair, the model identifier for the placement scenario after repair, the action combination identifier, a comparison of the placement confidence coefficient before and after, and a comparison of the uncertainty before and after.

[0121] Step S4 constructs a set of atomic actions within the local area defined by the minimum conflict breakpoint, including basic micro-movement, module orientation adjustment, crane station position translation, and temporary passage clearance. It defines the action amplitude boundary and availability conditions for the atomic actions, generates repair candidates based on the atomic action combination, and repeats step S3 to obtain the placement confidence coefficient and evidence chain. At the same time, it calculates the cost index and selects the final solution accordingly. It outputs the path key nodes, atomic action list, and evidence chain and writes them into the model change record for verification and record keeping.

[0122] Specifically, the above are merely preferred embodiments of this application and are not intended to limit this application.

[0123] The preset thresholds or preset parameters can be pre-calibrated through offline simulation testing or set to fixed values ​​according to on-site operating procedures.

[0124] In the description of this specification, references to terms such as "an embodiment," "example," and "specific example" indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0125] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for optimizing equipment layout in power transmission and transformation engineering design, characterized in that, Including the following steps: S1: Read the 3D model of the module equipment, extract the shape envelope, define the entrance unloading point and the foundation landing area, convert the site object into an obstacle body, and write the turning no-entry zone, the clearance constraint body and the allowable base pressure map to obtain the landing scene model. S2: The placement process is discretized into a finite state sequence of transportation, lifting, rotation and placement, which limits the attitude and position. The shape envelope generates the action sweep body and expands it to obtain the safe occupancy body. The obstacle body, the clearance constraint body and the load-bearing restricted area are gathered into a constraint set. S3: The candidate path set is obtained by searching the attitude connectivity from the unloading point to the target landing point. The conflict between the action sweep body and the constraint set is detected, and the conflict segment index and the trigger constraint object identifier are recorded to form the minimum conflict breakpoint marker. The minimum signed gap between the safe occupants and the obstacle sets is sampled along the candidate path set and resampled after stage alignment. The gap spectrum tail ratio is calculated based on the low gap tail segment. Extract the key working condition set from the candidate path set, calculate the equivalent base pressure of the outriggers and align the allowable base pressure map to obtain the base pressure ratio. Input the gap spectrum tail ratio and the base pressure ratio into the uncertainty inferrer with monotonic constraints to obtain the landing confidence coefficient and uncertainty. Output the minimum conflict breakpoint evidence chain and prune and sort it according to the landing confidence coefficient and uncertainty. S4: Construct a set of atomic actions in the local area based on the minimum conflict breakpoint. Define the action amplitude boundary and implementation conditions for each type of atomic action and combine them to generate repair candidates. Apply the repair candidates to the placement scene model and repeat step S3. At the same time, combine the cost index to complete the sorting and filtering, output the path key nodes, atomic action list and evidence chain, and write them into the model change record.

2. The equipment layout optimization method in power transmission and transformation engineering design according to claim 1, characterized in that, Step S1 includes: Read the 3D geometric model of the module equipment, extract the set of lifting points and the set of ports, calculate the envelope of the module shape and record the envelope orientation and the center of the envelope, analyze the constraints of transportation and hoisting methods to generate the set of allowable attitudes and the set of attitude switching conditions, establish a unified engineering coordinate benchmark, and obtain the placement scene model.

3. The equipment layout optimization method in power transmission and transformation engineering design according to claim 2, characterized in that, Step S1 also includes: The site enclosure boundary, existing framework, walls, roads, gates, and temporary facilities are converted into a set of obstacles and unified to a unified engineering coordinate benchmark. The entrance unloading point, target foundation placement area, transportation channel, crane position area, turning no-entry zone, and overhead clearance constraint are defined to form an allowable base pressure map and write it into the placement scene model.

4. The equipment layout optimization method in power transmission and transformation engineering design according to claim 3, characterized in that, Step S2 includes: Based on the placement scenario model, the placement process is discretized into a transportation section, a lifting section, a turning section, and a placement section. A set of state nodes containing working condition labels, attitude identifiers, and position identifiers is constructed, and allowable transfer boundaries are defined.

5. The equipment layout optimization method in power transmission and transformation engineering design according to claim 4, characterized in that, Step S2 also includes: Based on the set of state nodes, a key frame sequence consisting of construction method action nodes and geometric event nodes is generated. The action sweep volume is formed by using the module shape envelope and the lifting device shape envelope and then expanded to obtain the safety occupancy volume. The obstacle body set, the passage constraint body set, the turning no-entry zone, the overhead clearance constraint body and the load-bearing no-entry zone are solidified to form a constraint set and output the conflict determination interface.

6. The equipment layout optimization method in power transmission and transformation engineering design according to claim 5, characterized in that, Step S3 includes: A connectivity search is performed on the set of state nodes and the allowed transition boundary to generate a set of candidate paths. The safe occupancy body is called segment by segment according to the set of candidate paths, and conflict detection is performed with the constraint set. The conflict segment index and the trigger constraint object identifier are recorded to form the minimum conflict breakpoint marker.

7. The equipment layout optimization method in power transmission and transformation engineering design according to claim 6, characterized in that, Step S3 also includes: The minimum signed gap is sampled along the candidate path set to obtain the safe occupancy body and the obstacle body set, generating a gap sequence with working condition labels. Intra-stage consistency elimination and stage alignment resampling are performed to obtain the stage-aligned gap sequence. The gap spectrum tail ratio is calculated based on the low gap tail segment of the stage-aligned gap sequence.

8. The equipment layout optimization method in power transmission and transformation engineering design according to claim 7, characterized in that, Step S3 also includes: Extract the key working condition set from the candidate path set, calculate the equivalent base pressure of the outriggers and align the allowable base pressure map to obtain the base pressure residual ratio. Input the gap spectrum tail ratio and the base pressure residual ratio into the uncertainty inferrer with monotonic constraints to obtain the landing confidence coefficient and uncertainty. Perform candidate path pruning and sorting based on the landing confidence coefficient and uncertainty and output the evidence chain.

9. The equipment layout optimization method in power transmission and transformation engineering design according to claim 8, characterized in that, Step S4 includes: Based on the minimum conflict breakpoint, construct a set of atomic actions in the local area, define the action amplitude boundary and implementation conditions for each type of atomic action, and form a repair candidate generation rule table.

10. The equipment layout optimization method in power transmission and transformation engineering design according to claim 9, characterized in that, Step S4 also includes: A set of repair candidates is formed by combining atomic actions according to the repair candidate generation rule table. For each repair candidate, the recalculation is performed in the placement scenario model and step S3 is reused to obtain the placement confidence coefficient and evidence chain. At the same time, the cost index is combined to complete the sorting and screening, output the path key nodes, atomic action list and evidence chain, and write them into the model change record.

Images