Intelligent generation method and system for foundation pit construction scheme based on working condition constraint deduction

Abstract

This invention discloses an intelligent generation method and system for foundation pit construction schemes based on working condition constraint deduction. Through foundation pit engineering data acquisition and analysis, construction of a foundation pit engineering semantic object model, construction of an engineering constraint graph, retrieval and adaptation verification of foundation construction scheme templates, deduction of construction working condition chains, parameter collaborative optimization and generation of candidate foundation pit construction schemes, generation of complete sets of construction schemes, consistency verification, and final scheme output, the invention achieves intelligent and rapid generation of foundation pit construction schemes, improves the accuracy of scheme adaptation to actual foundation pit working conditions, reduces labor costs, and ensures foundation pit construction safety and engineering quality. It solves the problems of low efficiency, insufficient personalization, low matching degree with actual working conditions, and excessive reliance on human experience in existing foundation pit construction scheme generation technologies.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence, engineering knowledge graphs, and construction organization and safety control in foundation pit engineering. Specifically, it relates to an intelligent generation method, system, electronic device, and computer-readable storage medium for foundation pit construction schemes based on engineering semantic object models, engineering constraint graphs, and construction condition chain deduction. Background Technology

[0002] In construction engineering, municipal engineering, and underground space development, foundation pit construction is a critical stage characterized by high risk, wide impact, and strong process coupling. Foundation pit construction plans typically require comprehensive consideration of numerous factors, including the foundation pit's geometric dimensions, support structure type, engineering geological and hydrogeological conditions, distribution of surrounding buildings and underground pipelines, construction site conditions, machinery and equipment resources, construction period, safety level, and the company's construction methods. The quality of the plan directly affects the safety of the support structure, the effectiveness of groundwater control, environmental protection, construction progress, and project cost.

[0003] Currently, foundation pit construction plans mainly rely on manual preparation by engineering technicians. This manual approach requires repeated review of design drawings, engineering survey reports, standards and specifications, company construction methods, and similar historical plans. Based on this, sections on project overview, construction deployment, retaining wall construction, dewatering, earthwork excavation, monitoring and early warning, quality and safety measures, emergency plans, and acceptance requirements are organized. For projects in soft soil areas, with high groundwater levels, in sensitive surrounding environments, or with irregular foundation pit layouts, plan preparation often requires multiple rounds of adjustments, resulting in low efficiency, heavy reliance on experience, and inconsistencies in parameters between different plans.

[0004] Existing automated or intelligent tools mainly employ fixed template filling, keyword retrieval, standard clause matching, or compliance review of prepared plans. While these methods can improve text generation or review efficiency to some extent, they still have the following shortcomings: First, the systems primarily process construction plan texts or templates, lacking unified modeling of foundation pit geometry, support structures, groundwater control, surrounding environment, risk sources, and construction condition chains, making it difficult to understand foundation pit construction conditions from an engineering perspective; Second, existing methods often directly call templates based on similar cases, lacking automatic deduction of the dependencies between construction conditions such as "retaining wall construction—water-stop curtain—dewatering construction—test pumping—layered excavation—support construction—monitoring and early warning—emergency response"; Third, support, dewatering... Water, excavation, and monitoring are usually generated or verified separately, lacking a coupled optimization mechanism. This can easily lead to problems such as conflicts between dewatering wells and supports, ramps, main structures, or construction roads, as well as monitoring points not covering high-risk areas. Fourth, there is a lack of a unified parameter pool and consistency verification mechanism among the main body of the plan, technical parameter tables, construction process, dewatering well layout tables, monitoring point layout tables, calculation sheets, and attached drawing catalogs. A change in one parameter can easily cause multiple chapters to become out of sync. Fifth, existing intelligent review technologies focus on identifying and scoring problems in existing plans, but cannot perform partial reconstruction of the working condition chain and linkage correction of risk measures during the plan generation stage. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies, such as low efficiency in generating foundation pit construction plans, lack of personalization, low matching degree with actual working conditions, and excessive reliance on human experience. It provides an intelligent generation method and system for foundation pit construction plans based on working condition constraint deduction, realizing the intelligent and rapid generation of foundation pit construction plans, improving the accuracy of the plan's adaptation to the actual working conditions of the foundation pit, reducing labor costs, and ensuring the safety and quality of foundation pit construction.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] A method for intelligently generating foundation pit construction schemes based on working condition constraint deduction, comprising:

[0008] Step S1: Obtaining and analyzing foundation pit engineering data

[0009] Used for acquiring foundation pit engineering data, the data undergoes format recognition, data cleaning, and validity checks to form a set of data to be parsed, and the foundation pit engineering data is parsed to achieve engineering object identification;

[0010] Step S2: Construct the semantic object model for the foundation pit project.

[0011] The identified engineering objects are uniformly encoded and constructed into a semantic object model for foundation pit engineering, so that the subsequent scheme generation process is transformed from being driven by a single text to being driven by engineering objects.

[0012] Step S3: Construct the engineering constraint diagram

[0013] An engineering constraint map will be constructed based on engineering constraints formed by standards, design requirements, construction techniques, historical cases, and risk control rules.

[0014] Step S4: Retrieve the basic construction plan template and perform adaptation verification.

[0015] Based on the semantic object model of foundation pit engineering, foundation construction scheme templates are retrieved from the foundation pit construction scheme knowledge base;

[0016] Step S5: Derive the construction condition chain

[0017] Based on the semantic object model of foundation pit engineering, the template of foundation construction scheme and the engineering constraint diagram, the construction condition chain is automatically generated.

[0018] Step S6: Co-optimize parameters and generate candidate foundation pit construction schemes.

[0019] By optimizing the objective function, the parameters of the support structure, groundwater control, earthwork excavation, monitoring layout, and construction space conflict are used as coupling variables for joint optimization to generate candidate foundation pit construction schemes.

[0020] Step S7: Complete generation of construction plan

[0021] Based on the construction condition chain and optimized parameters, a special construction plan for the foundation pit is generated.

[0022] Step S8: Consistency Check and Final Solution Output

[0023] The parameters in the foundation pit construction plan are checked for consistency, including support structure conflict check, spatial conflict check, process logic check, dewatering effect check, monitoring coverage check, and emergency measure integrity check. If the check fails, the source of the conflict is located, the conflict parameters are adjusted according to the data source priority and engineering constraints, and then the process returns to step S1 until the output conditions are met or the maximum number of iterations is reached, thereby generating the final foundation pit construction plan.

[0024] Preferably, in step S1, the foundation pit engineering data includes drawings, geological survey data, surrounding environment data, standards and specifications, construction technology, design requirements, risk control rules, and historical cases. The foundation pit engineering data undergoes format recognition, data cleaning, and validity checks to form a set of data to be parsed. A natural language processing model is used to perform semantic recognition on the foundation pit engineering data. Line objects, block objects, layer names, text annotations, and dimension annotations are extracted from the drawings. Soil layer names, layer thickness, unit weight, cohesion, internal friction angle, permeability coefficient, groundwater level, and adverse geological conditions are extracted from the geological survey data. Buildings, roads, pipelines, rivers, and their spatial relationships with the foundation pit boundary are extracted from the surrounding environment data. The following engineering objects are obtained: foundation pit geometry, support structure, groundwater control, surrounding environment, construction resources, risk sources, and construction condition chain.

[0025] Preferably, in step S2, the semantic object model of the foundation pit engineering is represented as: P={G,S,W,E,C,R,T}, where G represents the foundation pit geometry, including the foundation pit boundary, excavation depth, excavation area, zone boundary, and exit location; S represents the support structure, including retaining piles, diaphragm walls, sheet piles, capping beams, supports, anchor cables, and cutoff walls; W represents groundwater control, including groundwater level, permeability coefficient, dewatering target water level, dewatering wells, and observation wells; E represents the surrounding environment, including existing buildings, underground pipelines, roads, rivers, and adjacent structures; C represents construction resources, including construction machinery, labor, materials, site conditions, and schedule requirements; R represents risk sources, including support deformation, pit bottom heave, quicksand, piping, surrounding settlement, pipeline damage, and rainstorm backflow; and T represents the construction condition chain, including retaining construction, dewatering construction, layered excavation, support construction, monitoring and early warning, and emergency response.

[0026] Preferably, in step S3, the nodes in the engineering constraint diagram include engineering object nodes, construction process nodes, specification clause nodes, risk event nodes, control measure nodes, monitoring project nodes, acceptance condition nodes, and emergency measure nodes; the relationships between nodes include "applicable to", "preceding", "subsequent", "constraint", "trigger", "conflict", "corresponding handling", and "requires verification".

[0027] The engineering constraints include constraints from standards and specifications, design requirements, construction techniques, historical cases, and risk control.

[0028] Preferably, in step S4, the foundation pit construction scheme knowledge base includes foundation construction scheme templates for different geological conditions, different support forms, different groundwater control methods, and different safety levels; the template matching score is calculated by comprehensively considering semantic similarity, rule matching degree, and historical case adaptability, and templates that have obvious conflicts with design requirements and construction technology are eliminated.

[0029] Preferably, in step S5, the construction condition chain is T=(N,E), where T is the construction condition chain, N is the set of construction condition nodes, and E is the dependency relationship between construction condition nodes.

[0030] Each construction condition node is represented as: N i =(O i M i ,P i ,R i Q i In the formula, N i For construction condition nodes, O i For the construction object, M i For construction methods, P i For construction parameters, R i For a set of risk events, Q i For quality and safety control requirements;

[0031] The generation of the construction condition chain satisfies the following optimization objectives:

[0032] ,in, For the optimal construction condition chain, For resource consumption, For construction risks, To constrain violations and penalties, The deviation from the project schedule is represented by α, β, γ, and δ, which are weighting coefficients. In the formula, T represents the construction condition chain, m is the total number of engineering constraints, i is the sequence number of engineering constraints greater than or equal to 1 and less than or equal to m, i is a natural number, and g i (T) is the construction condition chain T, used as input to determine whether the i-th engineering constraint is satisfied, λ i The penalty coefficient is the one corresponding to the i-th engineering constraint. When the construction condition chain violates the constraints of "support before excavation", "excavation can only be carried out after the water level reaches the target level", or "the difference in excavation height between adjacent zones shall not exceed the limit", the penalty value is automatically increased and the construction sequence is readjusted.

[0033] The work condition chain includes, but is not limited to, the following work condition nodes: construction preparation, surveying and setting out, retaining structure construction, water-stop curtain construction, capping beam construction, dewatering well construction, test pumping, layered and segmented excavation, support or anchor cable construction, monitoring and early warning, pit bottom acceptance, subbase construction, underground structure construction, support replacement and dismantling, backfilling and emergency response.

[0034] Preferably, in step S6, the optimization objective function is: In the formula, X* is the optimal coupling parameter solution that minimizes the total loss, and L... sTo mitigate safety losses, L w Losses due to insufficient groundwater control, L e For excavation disturbance losses, L m To monitor coverage loss, L c Losses due to space conflict.

[0035] Preferably, in step S6, the optimization objective function is obtained through a reinforcement learning model, and the reinforcement learning model has a built-in rule for calculating the quantization of component losses, using the support parameter set X. s Groundwater control parameter set X w Excavation parameter set X e Monitoring parameter set X m Spatial conflict parameter set X c Using these as input variables, and combining the foundation pit specification thresholds, real-time monitoring conditions of the foundation pit, and engineering constraints of the surrounding environment, a penalty-type loss function is constructed, and the support safety loss L is calculated sequentially. s Insufficient groundwater control resulted in losses of L. w Excavation disturbance loss L e Insufficient monitoring coverage leads to losses (L) m Space conflict loss L c When construction parameters exceed the safety control standard range, an exponential penalty mechanism is used to amplify the corresponding sub-item loss value.

[0036] Among them, X s This is a set of support parameters, including support structure type, pile length, pile diameter, pile spacing, number of support rows, and support axial force; X w This is a set of groundwater control parameters, including precipitation method, number of wells, well spacing, well depth, pumping flow rate, and target water level; X e This is a set of excavation parameters, including layer thickness, segment length, excavation sequence, excavation speed, and excavation path; X m This is a set of monitoring parameters, including monitoring point locations, monitoring items, monitoring frequency, alarm values, and control values; X c This is a set of spatial conflict parameters, including site spatial coordination parameters such as the layout of construction access roads, material storage areas, machinery operating radius, temporary support arrangement, and safety distance between pipelines and the main structure. The complete set of decision variables X={X...} is used. s ,X w ,X e ,X m ,X c The purpose is to jointly optimize the support structure parameters, groundwater control parameters, earthwork excavation parameters, monitoring layout parameters, and site spatial layout parameters as coupled variables.

[0037] As can be seen above, each sub-item of loss uses a corresponding subset of construction parameters as input, and combines the limits of the foundation pit specifications, real-time on-site conditions, and surrounding environmental constraints to construct a quantitative penalty calculation rule: based on the support parameter set X s The support deformation and stability coefficient are calculated and compared with the safety threshold to obtain the support safety loss Ls; based on the groundwater control parameter set X w Simulated water level drawdown and surface subsidence, compared with the allowable deformation standards for buildings and structures, yielded the loss L due to insufficient groundwater control. w Based on the excavation parameter set X e Calculate the soil unloading disturbance range and retaining wall lateral displacement, and obtain the excavation disturbance loss L based on environmental control indicators. e Based on monitoring parameter set X m The assessment of the completeness of monitoring point coverage at risk sections and the rationality of the early warning mechanism resulted in losses due to insufficient monitoring coverage. m Based on the spatial arrangement parameter set X c The spatial conflict loss L is obtained by calculating the safety clearance deviations and interference frequencies of machinery, loads, temporary components and pipelines, and the main structure. c The more each parameter deviates from the reasonable control range and the lower the safety reserve, the greater the corresponding penalty value for the sub-item loss. The total loss is obtained by summing the five sub-item losses. The reinforcement learning model continuously iterates and adjusts the entire set of coupling parameters X and repeats the above loss calculation process to finally find the optimal coupling parameter X* with the minimum total loss.

[0038] Preferably, in step S6, to achieve coordinated arrangement of dewatering wells and monitoring points inside and outside the foundation pit, during the generation of candidate foundation pit construction schemes, not only are boundary control wells generated along the foundation pit boundary, but also in-pit dewatering wells, risk enhancement wells, and groundwater observation wells are generated based on the excavation area inside the foundation pit; the foundation pit boundary and internal area are discretized into a set of candidate well points, and well points that conflict with the support structure, excavation ramp, main structure, construction road, and mechanical operation range are eliminated, and the number of well points, well depth, well spacing, filter pipe length, and pumping control requirements are determined according to the water level control target; the arrangement of monitoring points is linked with the arrangement of dewatering wells, and the monitoring frequency of groundwater level, support displacement, and surrounding settlement is automatically increased in areas with insufficient groundwater control or high risk level of the surrounding environment.

[0039] When conflicts and avoidances occur between the drainage wells, risk enhancement wells, observation wells, support systems, excavation ramps, and main structures within the pit, the pit boundary and internal area are discretized into a set of candidate well points: C h ={c1,c2,…,c n}, where C h Let n be the set of candidate well points, and c be the total number of candidate well points. i =(x i ,y i ,z iLet d(c) represent the coordinates of the i-th candidate well point, where i is a natural number greater than or equal to 1 and less than or equal to n; the candidate well point needs to satisfy the following spatial avoidance condition: d(c) i ,S)>d s d(c i ,Road)>d r d(c i, Structure)>d b In the formula, d(c i d(ci,Road) represents the minimum distance between the candidate well point and the supporting structure, and ds is the support avoidance distance; d(ci,Road) represents the minimum distance between the candidate well point and the construction road or excavation ramp, d r d(c) is the distance to avoid obstacles on the road. i (,Structure) represents the minimum distance between the candidate well point and the main structural member, d b For structural clearance distance;

[0040] The objective function for the layout of dewatering wells is: In the formula, W * It is the optimal combination of precipitation well layout parameters, N w E represents the number of well points. w For precipitation energy consumption, U w C is a penalty item for inadequate water level control. w For construction conflict penalties, η1, η2, η3, and η4 are weighting coefficients; for insufficient water level control penalties, the following are also considered: In the formula, h j To predict the water level at the j-th control point, h t To control the target water level, m represents the total number of control points, and the layout of monitoring points is linked with the layout of dewatering wells.

[0041] Preferably, in step S7, the foundation pit construction plan includes the main text of the plan, a technical parameter table, a calculation summary, a dewatering well layout table, a monitoring point layout table, and a list of attached drawings.

[0042] Preferably, in step S8, a unified parameter pool is set up. The main text of the plan, the technical parameter table, the calculation summary, the dewatering well layout table, the monitoring point layout table, and the attached drawing index all retrieve data from the unified parameter pool. When a parameter changes, the relevant parts of the plan are updated synchronously. The consistency loss of parameters of the same type is expressed as: In the formula, n represents the total number of parameters, and i and j are both natural numbers greater than or equal to 1 and less than or equal to n. and To unify the i-th and j-th parameters in the parameter pool, and Indicates the corresponding parameter and The parameter type, and Indicates the corresponding parameter and The parameter values, when If so, it is determined that there is a parameter conflict in the construction plan, and it is corrected according to the data source priority. The preset threshold;

[0043] Data source priority is set as follows: standard data takes precedence over geological survey data, geological survey data takes precedence over drawing data, and drawing data takes precedence over historical cases.

[0044] Preferably, in step S8, the output of the foundation pit special construction plan includes project overview, preparation basis, engineering geological and hydrogeological conditions, surrounding environment analysis, construction deployment, construction schedule, construction site layout, support structure construction method, groundwater control plan, earthwork excavation plan, monitoring and control plan, quality assurance measures, safety assurance measures, rainy season construction measures, emergency plan, acceptance requirements, machinery and equipment plan, labor plan, material plan, risk identification list, dewatering well layout table, monitoring point layout table, and attached drawing list.

[0045] This embodiment also provides an intelligent generation system for foundation pit construction schemes based on working condition constraint deduction, used to implement the intelligent generation method for foundation pit construction schemes based on working condition constraint deduction as described above. The generation system includes:

[0046] The foundation pit engineering data acquisition and parsing module is used to acquire foundation pit engineering data, perform format recognition, data cleaning and validity checks on the foundation pit engineering data, form a set of data to be parsed, and parse the foundation pit engineering data to realize the identification of engineering objects;

[0047] A semantic object model module for foundation pit engineering is constructed to uniformly encode the identified engineering objects and construct a semantic object model for foundation pit engineering.

[0048] The module for constructing engineering constraint maps is used to build engineering constraint maps based on standards, design requirements, construction techniques, historical cases, and risk control rules.

[0049] The module for retrieving and adapting basic construction scheme templates is used to retrieve basic construction scheme templates from the foundation pit construction scheme knowledge base based on the semantic object model of foundation pit engineering. It calculates the template matching score by comprehensively considering semantic similarity, rule matching degree, and historical case adaptation degree, and removes templates that have obvious conflicts with design requirements and construction technology.

[0050] The construction condition chain deduction module automatically generates construction condition chains based on the semantic object model of foundation pit engineering, foundation construction scheme templates, and engineering constraint diagrams.

[0051] The parameter co-optimization module is used to jointly optimize the support structure parameters, groundwater control parameters, earthwork excavation parameters, monitoring layout parameters, and construction space conflict parameters as coupling variables through the optimization objective function, and generate candidate foundation pit construction schemes.

[0052] The construction plan generation module is used to generate specialized construction plans for foundation pits based on the construction condition chain and optimized parameters.

[0053] The consistency verification and final scheme output module is used to verify the consistency of parameters in the foundation pit construction scheme, including support structure conflict verification, spatial conflict verification, process logic verification, dewatering effect verification, monitoring coverage verification, and emergency measure integrity verification. If the verification fails, the source of the conflict is located, and the conflicting parameters are adjusted according to the data source priority and engineering constraints until the output conditions are met or the maximum number of iterations is reached, thereby generating the final foundation pit construction scheme.

[0054] As can be seen from the above-disclosed technical solutions, the beneficial effects of the present invention compared with the prior art are as follows:

[0055] As can be seen from the above technical solutions, the beneficial effects of the present invention are as follows:

[0056] (1) A semantic object model for foundation pit engineering is proposed, which encodes the foundation pit geometry, support structure, groundwater control, surrounding environment, construction resources, risk sources and construction condition chain in a unified manner, so as to realize the object-oriented expression of the construction scheme generation process.

[0057] (2) A method for deducing engineering constraint diagrams and construction condition chains is proposed, which transforms specifications, design requirements, construction technology, historical cases and risk control rules into calculable constraints, thereby realizing the automatic deduction of foundation pit construction procedures.

[0058] (3) A collaborative optimization method for support-dewatering-excavation-monitoring is proposed, which uses support parameters, groundwater control parameters, excavation parameters, monitoring parameters and spatial conflict parameters as coupling variables for joint optimization, avoiding parameter conflicts and risk omissions caused by the independent generation of each special measure in the traditional method.

[0059] (4) A method for the combined arrangement of drainage wells, boundary control wells, risk enhancement wells and observation wells in the pit is proposed. This method can automatically avoid conflict areas based on the internal space of the pit, the support system, the excavation ramp, the construction road and the location of the main structure, thereby improving the constructability of the dewatering scheme.

[0060] (5) A method for verifying and automatically correcting the consistency of construction plans is proposed. The main text of the construction plan, the technical parameter table, the layout table of dewatering wells, the layout table of monitoring points, the calculation sheet and the catalog of attached drawings are checked in a coordinated manner. When inconsistencies in parameters, conflicts in procedures or missing risk measures are found, the method is automatically corrected. Attached Figure Description

[0061] Figure 1 This is a flowchart illustrating the intelligent generation method for foundation pit construction schemes based on working condition constraints. Detailed Implementation

[0062] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The technical content and features of the present invention will be described in detail below with reference to the listed embodiments and the accompanying drawings. It should be noted that the drawings are all in a very simplified form and use non-precise proportions, and are only used to facilitate and clearly illustrate the embodiments of the present invention. For ease of description, the terms "upper" and "lower" used below are consistent with the upper and lower directions in the drawings, but this should not be construed as a limitation of the technical solution of the present invention.

[0063] Please see Figure 1 This embodiment discloses an intelligent generation method for foundation pit construction schemes based on working condition constraint deduction, including:

[0064] Step S1: Obtaining and analyzing foundation pit engineering data

[0065] Used for acquiring foundation pit engineering data, the data undergoes format recognition, data cleaning, and validity checks to form a set of data to be parsed, and the foundation pit engineering data is parsed to achieve engineering object identification;

[0066] Step S2: Construct the semantic object model for the foundation pit project.

[0067] The identified engineering objects are uniformly encoded and constructed into a semantic object model for foundation pit engineering, so that the subsequent scheme generation process is transformed from being driven by a single text to being driven by engineering objects.

[0068] Step S3: Construct the engineering constraint diagram

[0069] An engineering constraint map will be constructed based on engineering constraints formed by standards, design requirements, construction techniques, historical cases, and risk control rules.

[0070] Step S4: Retrieve the basic construction plan template and perform adaptation verification.

[0071] Based on the semantic object model of foundation pit engineering, foundation construction scheme templates are retrieved from the foundation pit construction scheme knowledge base;

[0072] Step S5: Derive the construction condition chain

[0073] Based on the semantic object model of foundation pit engineering, the template of foundation construction scheme and the engineering constraint diagram, the construction condition chain is automatically generated.

[0074] Step S6: Co-optimize parameters and generate candidate foundation pit construction schemes.

[0075] By optimizing the objective function, the parameters of the support structure, groundwater control, earthwork excavation, monitoring layout, and construction space conflict are used as coupling variables for joint optimization to generate candidate foundation pit construction schemes.

[0076] Step S7: Complete generation of construction plan

[0077] Based on the construction condition chain and optimized parameters, a special construction plan for the foundation pit is generated.

[0078] Step S8: Consistency Check and Final Solution Output

[0079] The parameters in the foundation pit construction plan are checked for consistency, including support structure conflict check, spatial conflict check, process logic check, dewatering effect check, monitoring coverage check, and emergency measure integrity check. If the check fails, the source of the conflict is located, the conflict parameters are adjusted according to the data source priority and engineering constraints, and then the process returns to step S1 until the output conditions are met or the maximum number of iterations is reached, thereby generating the final foundation pit construction plan.

[0080] Preferably, in step S1, the foundation pit engineering data includes drawings, geological survey data, surrounding environment data, standards and specifications, construction technology, design requirements, risk control rules, and historical cases. The foundation pit engineering data undergoes format recognition, data cleaning, and validity checks to form a set of data to be parsed. A natural language processing model is used to perform semantic recognition on the foundation pit engineering data. Line objects, block objects, layer names, text annotations, and dimension annotations are extracted from the drawings. Soil layer names, layer thickness, unit weight, cohesion, internal friction angle, permeability coefficient, groundwater level, and adverse geological conditions are extracted from the geological survey data. Buildings, roads, pipelines, rivers, and their spatial relationships with the foundation pit boundary are extracted from the surrounding environment data. The following engineering objects are obtained: foundation pit geometry, support structure, groundwater control, surrounding environment, construction resources, risk sources, and construction condition chain.

[0081] Preferably, in step S2, the semantic object model of the foundation pit engineering is represented as: P={G,S,W,E,C,R,T}, where G represents the foundation pit geometry, including the foundation pit boundary, excavation depth, excavation area, zone boundary, and exit location; S represents the support structure, including retaining piles, diaphragm walls, sheet piles, capping beams, supports, anchor cables, and cutoff curtains; W represents groundwater control, including groundwater level, permeability coefficient, dewatering target water level, dewatering wells, and observation wells; E represents the surrounding environment, including existing buildings, underground pipelines, roads, rivers, and adjacent structures; C represents construction resources, including construction machinery, labor, materials, site conditions, and schedule requirements; R represents risk sources, including support deformation, pit bottom heave, quicksand, piping, surrounding settlement, pipeline damage, and rainstorm backflow; and T represents the construction condition chain, including retaining construction, dewatering construction, layered excavation, support construction, monitoring and early warning, and emergency response. Therefore, by uniformly encoding the foundation pit geometry, support structure, groundwater control, surrounding environment, construction resources, risk sources, and construction condition chain, this invention can realize the engineering object-oriented expression of the construction scheme generation process.

[0082] Preferably, in step S3, the engineering constraints include standard constraints, design requirement constraints, construction process constraints, historical case constraints, and risk control constraints. Specifically, standard constraints include a sliding safety factor ≥1.3 for primary foundation pit support; design requirement constraints include vertical spacing of supports not exceeding 4m; construction process constraints include supporting before excavation and layer thickness ≤3m; historical case constraints include a daily water level drop of no more than 0.5m for soft soil foundation pits; and risk control constraints include prohibiting heavy loading within 3m of the foundation pit boundary. As can be seen, this invention transforms standard constraints, design requirements, construction processes, historical cases, and risk control rules into calculable constraints, enabling the automatic deduction of foundation pit construction procedures.

[0083] The nodes in the engineering constraint diagram include engineering object nodes, construction procedure nodes, specification clause nodes, risk event nodes, control measure nodes, monitoring project nodes, acceptance condition nodes, and emergency measure nodes; the relationships between nodes include "applicable to", "preceding", "subsequent", "constraint", "trigger", "conflict", "corresponding handling", and "requires verification".

[0084] Preferably, in step S4, the foundation pit construction scheme knowledge base includes foundation construction scheme templates for different geological conditions, different support forms, different groundwater control methods, and different safety levels; the template matching score is calculated by comprehensively considering semantic similarity, rule matching degree, and historical case adaptability, and templates that have obvious conflicts with design requirements and construction technology are eliminated.

[0085] Preferably, in step S5, the construction condition chain is T=(N,E), where T is the construction condition chain, N is the set of construction condition nodes, and E is the dependency relationship between construction condition nodes.

[0086] Each construction condition node is represented as: N i =(O i M i ,P i ,R i Q i In the formula, N i For construction condition nodes, O i For the construction object, M i For construction methods, P i For construction parameters, R i For a set of risk events, Q i For quality and safety control requirements;

[0087] The generation of the construction condition chain satisfies the following optimization objectives:

[0088] ,in, For the optimal construction condition chain, For resource consumption, For construction risks, To constrain violations and penalties, The deviation from the project schedule is represented by α, β, γ, and δ, which are weighting coefficients with values ​​greater than 0. These coefficients are used to balance the numerical magnitudes of the four indicators with the optimization priority, and there are no fixed summation constraints on the weighting coefficients. In the formula, T represents the construction condition chain, m is the total number of engineering constraints, i is the sequence number of engineering constraints greater than or equal to 1 and less than or equal to m, i is a natural number, and g i (T) is the construction condition chain T, used as input to determine whether the i-th engineering constraint is satisfied, λ i The penalty coefficient is the one corresponding to the i-th engineering constraint. When the construction condition chain violates the constraints of "support before excavation", "excavation can only be carried out after the water level reaches the target level", or "the difference in excavation height between adjacent zones shall not exceed the limit", the penalty value is automatically increased and the construction sequence is readjusted.

[0089] The work condition chain includes, but is not limited to, the following work condition nodes: construction preparation, surveying and setting out, retaining structure construction, water-stop curtain construction, capping beam construction, dewatering well construction, test pumping, layered and segmented excavation, support or anchor cable construction, monitoring and early warning, pit bottom acceptance, subbase construction, underground structure construction, support replacement and dismantling, backfilling and emergency response.

[0090] Preferably, in step S6, the optimization objective function is: In the formula, X* is the optimal coupling parameter solution that minimizes the total loss, and L... s To mitigate safety losses, L w Losses due to insufficient groundwater control, L e For excavation disturbance losses, L m To monitor coverage loss, L c This invention addresses spatial conflict losses by using support parameters, groundwater control parameters, excavation parameters, monitoring parameters, and spatial conflict parameters as coupled variables for joint optimization. This avoids parameter conflicts and risk omissions caused by the independent generation of each specific measure in traditional methods.

[0091] Preferably, in step S6, the optimization objective function is obtained through a reinforcement learning model, which has a built-in rule for calculating the quantification of component losses, using the support parameter set Xs and the groundwater control parameter set X w Excavation parameter set Xe, monitoring parameter set X m Spatial conflict parameter set X c Using these as input variables, and combining the foundation pit specification thresholds, real-time monitoring conditions of the foundation pit, and engineering constraints of the surrounding environment, a penalty-type loss function is constructed, and the support safety loss L is calculated sequentially. s Insufficient groundwater control resulted in losses of L. w Excavation disturbance loss L e Insufficient monitoring coverage leads to losses (L) m Space conflict loss L c When construction parameters exceed the safety control standard range, an exponential penalty mechanism is used to amplify the corresponding sub-item loss value.

[0092] Among them, X s This is a set of support parameters, including support structure type, pile length, pile diameter, pile spacing, number of support rows, and support axial force; X w This is a set of groundwater control parameters, including precipitation method, number of wells, well spacing, well depth, pumping flow rate, and target water level; X e This is a set of excavation parameters, including layer thickness, segment length, excavation sequence, excavation speed, and excavation path; X m This is a set of monitoring parameters, including monitoring point locations, monitoring items, monitoring frequency, alarm values, and control values; X c This is a set of spatial conflict parameters, including site spatial coordination parameters such as the layout of construction access roads, material storage areas, machinery operating radius, temporary support arrangement, and safety distance between pipelines and the main structure. The complete set of decision variables X={X...} is used. s ,X w ,X e ,X m ,X cThe purpose is to jointly optimize the support structure parameters, groundwater control parameters, earthwork excavation parameters, monitoring layout parameters, and site spatial layout parameters as coupled variables.

[0093] As can be seen above, each sub-item of loss uses a corresponding subset of construction parameters as input, and combines the limits of the foundation pit specifications, real-time on-site conditions, and surrounding environmental constraints to construct a quantitative penalty calculation rule: based on the support parameter set X s The support deformation and stability coefficient are calculated and compared with the safety threshold to obtain the support safety loss Ls; based on the groundwater control parameter set X w Simulated water level drawdown and surface subsidence, compared with the allowable deformation standards for buildings and structures, yielded the loss L due to insufficient groundwater control. w Based on the excavation parameter set X e Calculate the soil unloading disturbance range and retaining wall lateral displacement, and obtain the excavation disturbance loss L based on environmental control indicators. e Based on monitoring parameter set X m The assessment of the completeness of monitoring point coverage at risk sections and the rationality of the early warning mechanism resulted in losses due to insufficient monitoring coverage. m Based on the spatial arrangement parameter set X c The spatial conflict loss L is obtained by calculating the safety clearance deviations and interference frequencies of machinery, loads, temporary components and pipelines, and the main structure. c The greater the deviation of various parameters from the reasonable control range and the lower the safety margin, the larger the corresponding penalty value for each component loss. The total loss is obtained by summing the five component losses. The reinforcement learning model continuously iterates and adjusts the entire set of coupling parameters X and repeats the above loss calculation process to finally find the optimal coupling parameters X that minimize the total loss. * .

[0094] Preferably, in step S6, to achieve coordinated arrangement of dewatering wells and monitoring points inside and outside the foundation pit, during the generation of candidate foundation pit construction schemes, not only are boundary control wells generated along the foundation pit boundary, but also in-pit dewatering wells, risk enhancement wells, and groundwater observation wells are generated based on the excavation area inside the foundation pit; the foundation pit boundary and internal area are discretized into a set of candidate well points, and well points that conflict with the support structure, excavation ramp, main structure, construction road, and mechanical operation range are eliminated, and the number of well points, well depth, well spacing, filter pipe length, and pumping control requirements are determined according to the water level control target; the arrangement of monitoring points is linked with the arrangement of dewatering wells, and the monitoring frequency of groundwater level, support displacement, and surrounding settlement is automatically increased in areas with insufficient groundwater control or high risk level of the surrounding environment.

[0095] When conflicts and avoidances occur between the drainage wells, risk enhancement wells, observation wells, support systems, excavation ramps, and main structures within the pit, the pit boundary and internal area are discretized into a set of candidate well points: C h ={c1,c2,…,c n}, where Ch Let n be the set of candidate well points, and c be the total number of candidate well points. i =(x i ,y i ,z i Let d(c) represent the coordinates of the i-th candidate well point, where i is a natural number greater than or equal to 1 and less than or equal to n; the candidate well point needs to satisfy the following spatial avoidance condition: d(c) i ,S)>d s d(c i ,Road)>d r d(c i, Structure)>d b In the formula, d(c i d(ci,Road) represents the minimum distance between the candidate well point and the supporting structure, and ds is the support avoidance distance; d(ci,Road) represents the minimum distance between the candidate well point and the construction road or excavation ramp, d r d(c) is the distance to avoid obstacles on the road. i (,Structure) represents the minimum distance between the candidate well point and the main structural member, d b For structural clearance distance;

[0096] The objective function for the layout of dewatering wells is: In the formula, W * It is the optimal combination of precipitation well layout parameters, N w E represents the number of well points. w For precipitation energy consumption, U w C is a penalty item for inadequate water level control. w For construction conflict penalties, η1, η2, η3, and η4 are weighting coefficients; for insufficient water level control penalties, the following are also considered: In the formula, h j To predict the water level at the j-th control point, h t To control the target water level, where m represents the total number of control points, the layout of monitoring points is linked to the layout of dewatering wells. This invention, through a combined layout method of in-pit dewatering wells, boundary control wells, risk reinforcement wells, and observation wells, can automatically avoid conflict areas based on the internal space of the foundation pit, the support system, the excavation ramp, the construction road, and the location of the main structure, thereby improving the constructability of the dewatering scheme.

[0097] Preferably, in step S7, the foundation pit construction plan includes the main text of the plan, a technical parameter table, a calculation summary, a dewatering well layout table, a monitoring point layout table, and a list of attached drawings.

[0098] Preferably, in step S8, a unified parameter pool is set up. The main text of the scheme, the technical parameter table, the calculation summary, the dewatering well layout table, the monitoring point layout table, and the attached drawing index all retrieve data from the unified parameter pool. When a parameter changes, the relevant parts of the scheme are updated synchronously. The consistency loss of the same type of parameter is expressed as: In the formula, This represents the consistency loss for parameters of the same type; n represents the total number of parameters, and i and j are both natural numbers greater than or equal to 1 and less than or equal to n. and To unify the i-th and j-th parameters in the parameter pool, and Indicates the corresponding parameter and The parameter type, and Indicates the corresponding parameter and The parameter values, when If so, it is determined that there is a parameter conflict in the construction plan, and it is corrected according to the data source priority. This is a preset threshold.

[0099] I( = ) is an indicator function, defined by the following rules:

[0100] When the condition inside the parentheses is true (a) i a j If they are parameters of the same type, such as both being "excavation depth of the foundation pit"), I(True)=1;

[0101] When the condition within the parentheses is not true (a i a j The parameter types are different: one is the excavation depth, and the other is the thickness of the diaphragm wall: I(False)=0.

[0102] The meaning is to double-sum iterate through all pairwise parameter pairs (a) in the parameter pool. i ,a j Only when two parameters belong to the same parameter type (I=1) will the absolute value of their numerical difference be included in the loss; for parameters of different types (I=0), this item is directly cleared to zero and does not participate in the loss accumulation. For example:

[0103] a i =The excavation depth on the drawing is 8.5m, a j =Text excavation depth 9.0m: Same type, I=1, contribution deviation |8.5-9.0|=0.5;

[0104] a i =Excavation depth, a j =Support pile diameter: For different types, I=0, and the loss is not included.

[0105] The data source priority is set as follows: standard specifications take precedence over geological survey data, geological survey data takes precedence over drawing data, and drawing data takes precedence over historical cases. Therefore, this invention proposes a method for verifying and automatically correcting the consistency of construction plans, linking the main text of the construction plan, technical parameter tables, dewatering well layout tables, monitoring point layout tables, calculation sheets, and attached drawing catalogs for verification. It automatically corrects inconsistencies in parameters, conflicts in procedures, or missing risk measures when they are found.

[0106] Preferably, in step S8, the output of the foundation pit special construction plan includes project overview, preparation basis, engineering geological and hydrogeological conditions, surrounding environment analysis, construction deployment, construction schedule, construction site layout, support structure construction method, groundwater control plan, earthwork excavation plan, monitoring and control plan, quality assurance measures, safety assurance measures, rainy season construction measures, emergency plan, acceptance requirements, machinery and equipment plan, labor plan, material plan, risk identification list, dewatering well layout table, monitoring point layout table, and attached drawing list.

[0107] This embodiment also provides an intelligent generation system for foundation pit construction schemes based on working condition constraint deduction, used to implement the intelligent generation method for foundation pit construction schemes based on working condition constraint deduction as described above. The generation system includes:

[0108] The foundation pit engineering data acquisition and parsing module is used to acquire foundation pit engineering data, perform format recognition, data cleaning and validity checks on the foundation pit engineering data, form a set of data to be parsed, and parse the foundation pit engineering data to realize the identification of engineering objects;

[0109] A semantic object model module for foundation pit engineering is constructed to uniformly encode the identified engineering objects and construct a semantic object model for foundation pit engineering.

[0110] The module for constructing engineering constraint maps is used to build engineering constraint maps based on standards, design requirements, construction techniques, historical cases, and risk control rules.

[0111] The module for retrieving and adapting basic construction scheme templates is used to retrieve basic construction scheme templates from the foundation pit construction scheme knowledge base based on the semantic object model of foundation pit engineering. It calculates the template matching score by comprehensively considering semantic similarity, rule matching degree, and historical case adaptation degree, and removes templates that have obvious conflicts with design requirements and construction technology.

[0112] The construction condition chain deduction module automatically generates construction condition chains based on the semantic object model of foundation pit engineering, foundation construction scheme templates, and engineering constraint diagrams.

[0113] The parameter co-optimization module is used to jointly optimize the support structure parameters, groundwater control parameters, earthwork excavation parameters, monitoring layout parameters, and construction space conflict parameters as coupling variables through the optimization objective function, and generate candidate foundation pit construction schemes.

[0114] The construction plan generation module is used to generate specialized construction plans for foundation pits based on the construction condition chain and optimized parameters.

[0115] The consistency verification and final scheme output module is used to verify the consistency of parameters in the foundation pit construction scheme, including support structure conflict verification, spatial conflict verification, process logic verification, dewatering effect verification, monitoring coverage verification, and emergency measure integrity verification. If the verification fails, the source of the conflict is located, and the conflict parameters are adjusted according to the data source priority and engineering constraints. The process is then iterated back until the output conditions are met or the maximum number of iterations is reached, thereby generating the final foundation pit construction scheme.

[0116] In summary, the beneficial effects of the present invention are as follows:

[0117] (1) A semantic object model for foundation pit engineering is proposed, which encodes the foundation pit geometry, support structure, groundwater control, surrounding environment, construction resources, risk sources and construction condition chain in a unified manner, so as to realize the object-oriented expression of the construction scheme generation process.

[0118] (2) A method for deducing engineering constraint diagrams and construction condition chains is proposed, which transforms specifications, design requirements, construction technology, historical cases and risk control rules into calculable constraints, thereby realizing the automatic deduction of foundation pit construction procedures.

[0119] (3) A collaborative optimization method for support-dewatering-excavation-monitoring is proposed, which uses support parameters, groundwater control parameters, excavation parameters, monitoring parameters and spatial conflict parameters as coupling variables for joint optimization, avoiding parameter conflicts and risk omissions caused by the independent generation of each special measure in the traditional method.

[0120] (4) A method for the combined arrangement of drainage wells, boundary control wells, risk enhancement wells and observation wells in the pit is proposed. This method can automatically avoid conflict areas based on the internal space of the pit, the support system, the excavation ramp, the construction road and the location of the main structure, thereby improving the constructability of the dewatering scheme.

[0121] (5) A method for consistency verification and automatic correction of construction plans is proposed. This method performs a linked verification of the main text of the construction plan, technical parameter tables, dewatering well layout tables, monitoring point layout tables, calculation sheets, and a list of attached drawings. It automatically corrects inconsistencies in parameters, conflicts in procedures, or missing risk measures. The above description is merely a description of a preferred embodiment of the present invention and is not intended to limit the scope of the invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure are within the scope of protection of the claims.

Claims

1. A method for intelligently generating foundation pit construction schemes based on working condition constraint deduction, characterized in that, include: Step S1: Obtaining and analyzing foundation pit engineering data Used for acquiring foundation pit engineering data, the data undergoes format recognition, data cleaning, and validity checks to form a set of data to be parsed, and the foundation pit engineering data is parsed to achieve engineering object identification; Step S2: Construct the semantic object model for the foundation pit project. The identified engineering objects are uniformly encoded and constructed into a semantic object model for foundation pit engineering, so that the subsequent scheme generation process is transformed from being driven by a single text to being driven by engineering objects. Step S3: Construct the engineering constraint diagram An engineering constraint map will be constructed based on engineering constraints formed by standards, design requirements, construction techniques, historical cases, and risk control rules. Step S4: Retrieve the basic construction plan template and perform adaptation verification. Based on the semantic object model of foundation pit engineering, foundation construction scheme templates are retrieved from the foundation pit construction scheme knowledge base; Step S5: Derive the construction condition chain Based on the semantic object model of foundation pit engineering, the template of foundation construction scheme and the engineering constraint diagram, the construction condition chain is automatically generated. Step S6: Co-optimize parameters and generate candidate foundation pit construction schemes. By optimizing the objective function, the parameters of the support structure, groundwater control, earthwork excavation, monitoring layout, and construction space conflict are used as coupling variables for joint optimization to generate candidate foundation pit construction schemes. Step S7: Complete generation of construction plan Based on the construction condition chain and optimized parameters, a special construction plan for the foundation pit is generated. Step S8: Consistency Check and Final Solution Output The parameters in the foundation pit construction plan are checked for consistency, including the check for conflict of support structure, spatial conflict, process logic, dewatering effect, monitoring coverage and completeness of emergency measures. If the verification fails, the source of the conflict is located, the conflict parameters are adjusted according to the data source priority and engineering constraints, and then the process is backtracked and iterated. Then, the process is repeated until the output conditions are met or the maximum number of iterations is reached, thereby generating the final foundation pit construction plan.

2. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 1, characterized in that, In step S2, the semantic object model of the foundation pit engineering is represented as: P={G,S,W,E,C,R,T}, where G represents the foundation pit geometry, including the foundation pit boundary, excavation depth, excavation area, zone boundary, and exit location; S represents the support structure, including retaining piles, diaphragm walls, sheet piles, capping beams, supports, anchor cables, and cutoff curtains; W represents groundwater control, including groundwater level, permeability coefficient, dewatering target water level, dewatering wells, and observation wells; E represents the surrounding environment, including existing buildings, underground pipelines, roads, rivers, and adjacent structures; C represents construction resources, including construction machinery, labor, materials, site conditions, and schedule requirements; R represents risks, including support deformation, pit bottom heave, quicksand, piping, surrounding settlement, pipeline damage, and rainstorm backflow; and T represents the construction condition chain, including retaining construction, dewatering construction, layered excavation, support construction, monitoring and early warning, and emergency response.

3. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 1, characterized in that, In step S3, the engineering constraints include standard constraints, design requirement constraints, construction process constraints, historical case constraints, and risk control constraints; the nodes in the engineering constraint diagram include engineering object nodes, construction procedure nodes, standard clause nodes, risk event nodes, control measure nodes, monitoring item nodes, acceptance condition nodes, and emergency measure nodes; the relationships between nodes include "applicable to", "preceding", "subsequent", "constraint", "trigger", "conflict", "corresponding handling", and "requires verification".

4. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 1, characterized in that, In step S4, the foundation pit construction scheme knowledge base contains foundation construction scheme templates with different geological conditions, different support forms, different groundwater control methods, and different safety levels; the template matching score is calculated by comprehensively calculating semantic similarity, rule matching degree, and historical case adaptability, and templates that have obvious conflicts with design requirements and construction technology are eliminated.

5. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 1, characterized in that, In step S5, the construction condition chain is defined as T = (N, E), where T is the construction condition chain, N is the set of construction condition nodes, and E represents the dependencies between construction condition nodes. Each construction condition node is represented as: N i =(O i M i ,P i ,R i Q i In the formula, N i For construction condition nodes, O i For the construction object, M i For construction methods, P i For construction parameters, R i For a set of risk events, Q i For quality and safety control requirements; The generation of the construction condition chain satisfies the following optimization objectives: ,in, For the optimal construction condition chain, For resource consumption, For construction risks, To constrain violations and penalties, The deviation from the project schedule is represented by α, β, γ, and δ, which are weighting coefficients. In the formula, T represents the construction condition chain, m is the total number of engineering constraints, i is the sequence number of engineering constraints greater than or equal to 1 and less than or equal to m, and g i (T) is the construction condition chain T, used as input to determine whether the i-th engineering constraint is satisfied, λ i The penalty coefficient is the one corresponding to the i-th engineering constraint. When the construction condition chain violates the constraints of "support before excavation", "excavation can only be carried out after the water level reaches the target level", or "the difference in excavation height between adjacent zones shall not exceed the limit", the penalty value is automatically increased and the construction sequence is readjusted.

6. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 1, characterized in that, In step S6, the objective function for optimization is: In the formula, X* is the optimal coupling parameter solution that minimizes the total loss, and L... s To mitigate safety losses, L w Losses due to insufficient groundwater control, L e For excavation disturbance losses, L m To monitor coverage loss, L c Losses due to space conflict.

7. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 6, characterized in that, In step S6, the optimization objective function is obtained through a reinforcement learning model. The reinforcement learning model has built-in rules for calculating the quantification of component losses, using the support parameter set Xs and the groundwater control parameter set X. w Excavation parameter set Xe, monitoring parameter set X m Spatial conflict parameter set X c Using these as input variables, and combining the foundation pit specification thresholds, real-time monitoring conditions of the foundation pit, and engineering constraints of the surrounding environment, a penalty-type loss function is constructed, and the support safety loss L is calculated sequentially. s Insufficient groundwater control resulted in losses of L. w Excavation disturbance loss L e Insufficient monitoring coverage leads to losses (L) m Space conflict loss L c When construction parameters exceed the safety control standard range, an exponential penalty mechanism is used to amplify the corresponding sub-item loss value.

8. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 6, characterized in that, In step S6, to achieve coordinated layout of dewatering wells and monitoring points inside and outside the foundation pit, during the generation of candidate foundation pit construction schemes, not only are boundary control wells generated along the foundation pit boundary, but also in-pit dewatering wells, risk enhancement wells, and groundwater observation wells are generated based on the excavation area inside the foundation pit. The foundation pit boundary and internal area are discretized into a set of candidate well points, and well points that conflict with the support structure, excavation ramp, main structure, construction road, and mechanical operation range are eliminated. The number of well points, well depth, well spacing, filter pipe length, and pumping control requirements are determined according to the water level control target. The layout of monitoring points is linked with the layout of dewatering wells, and the monitoring frequency of groundwater level, support displacement, and surrounding settlement is automatically increased in areas with insufficient groundwater control or high risk levels of the surrounding environment. When conflicts and avoidances occur between the drainage wells, risk enhancement wells, observation wells, support systems, excavation ramps, and main structures within the pit, the pit boundary and internal area are discretized into a set of candidate well points: C h ={c1,c2,…,c n }, where C h It is the set of candidate well points, where n is the total number of candidate well points, and c i =(x i ,y i ,z i Let be the coordinates of the i-th candidate well point, where i is a natural number greater than or equal to 1 and less than or equal to n; Candidate well points need to meet the following spatial avoidance conditions: d(c i ,S)>d s d(c i ,Road)>d r d(c i, Structure)>d b In the formula, d(c i d(ci,Road) represents the minimum distance between the candidate well point and the supporting structure, and ds is the support avoidance distance; d(ci,Road) represents the minimum distance between the candidate well point and the construction road or excavation ramp, d r d(c) is the distance to avoid obstacles on the road. i (,Structure) represents the minimum distance between the candidate well point and the main structural member, d b For structural clearance distance; The objective function for the layout of dewatering wells is: In the formula, W * It is the optimal combination of precipitation well layout parameters, N w E represents the number of well points. w For precipitation energy consumption, U w C is a penalty item for inadequate water level control. w For construction conflict penalties, η1, η2, η3, and η4 are weighting coefficients; for insufficient water level control penalties, the following are also considered: In the formula, h j To predict the water level at the j-th control point, h t To control the target water level, m represents the total number of control points, and the layout of monitoring points is linked with the layout of dewatering wells.

9. The intelligent generation method for foundation pit construction scheme based on working condition constraint deduction as described in claim 1, characterized in that, In step S8, a unified parameter pool is set up. The main text of the plan, the technical parameter table, the calculation summary, the dewatering well layout table, the monitoring point layout table, and the attached figure catalog all retrieve data from the unified parameter pool. When a parameter changes, the relevant parts of the plan are updated synchronously. The consistency loss of the same type of parameter is represented as: In the formula, n represents the total number of parameters, and i and j are both natural numbers greater than or equal to 1 and less than or equal to n. and To unify the i-th and j-th parameters in the parameter pool, and Indicates the corresponding parameter and The parameter type, and Indicates the corresponding parameter and The parameter values, when If so, it is determined that there is a parameter conflict in the construction plan, and it is corrected according to the data source priority. The preset threshold is used; the data source priority is set as follows: standard data is higher than geological survey data, geological survey data is higher than drawing data, and drawing data is higher than historical cases.

10. An intelligent generation system for foundation pit construction schemes based on working condition constraint deduction, used to implement the intelligent generation method for foundation pit construction schemes based on working condition constraint deduction as described in any one of claims 1 to 9, wherein the generation system comprises: The foundation pit engineering data acquisition and parsing module is used to acquire foundation pit engineering data, perform format recognition, data cleaning and validity checks on the foundation pit engineering data, form a set of data to be parsed, and parse the foundation pit engineering data to realize the identification of engineering objects; A semantic object model module for foundation pit engineering is constructed to uniformly encode the identified engineering objects and construct a semantic object model for foundation pit engineering. The module for constructing engineering constraint maps is used to build engineering constraint maps based on standards, design requirements, construction techniques, historical cases, and risk control rules. The module for retrieving and adapting basic construction scheme templates is used to retrieve basic construction scheme templates from the foundation pit construction scheme knowledge base based on the semantic object model of foundation pit engineering. It calculates the template matching score by comprehensively considering semantic similarity, rule matching degree, and historical case adaptation degree, and removes templates that have obvious conflicts with design requirements and construction technology. The construction condition chain deduction module automatically generates construction condition chains based on the semantic object model of foundation pit engineering, foundation construction scheme templates, and engineering constraint diagrams. The parameter co-optimization module is used to jointly optimize the support structure parameters, groundwater control parameters, earthwork excavation parameters, monitoring layout parameters, and construction space conflict parameters as coupling variables through the optimization objective function, and generate candidate foundation pit construction schemes. The construction plan generation module is used to generate specialized construction plans for foundation pits based on the construction condition chain and optimized parameters. The consistency verification and final scheme output module is used to verify the consistency of parameters in the foundation pit construction scheme, including support structure conflict verification, spatial conflict verification, process logic verification, dewatering effect verification, monitoring coverage verification, and emergency measure integrity verification. If the verification fails, the source of the conflict is located, and the conflict parameters are adjusted according to the data source priority and engineering constraints. The process is then backtracked and iterated until the output conditions are met or the maximum number of iterations is reached, thereby generating the final foundation pit construction plan.

Images