A BIM-based whole-process intelligent management and control method and system for engineering supervision

Abstract

The application discloses a kind of based on BIM's engineering supervision whole process intelligent management and control method and system, and it is related to engineering supervision technical field.A kind of based on BIM's engineering supervision whole process intelligent management and control system, including have: engineering supervision prediction module, engineering supervision analysis module and engineering supervision decision module.The application is based on BIM model, engineering instruction semantic analysis and engineering energy consumption data constructs engineering supervision whole process intelligent management and control system, by plan construction behavior sequence construction, energy consumption benchmark fingerprint generation, equipment operation characteristic identification, actual construction behavior reconstruction and construction consistency deviation analysis, realize the automatic perception and intelligent supervision to construction site real job process;Continuously obtain dynamic state information in construction process, promptly find construction deviation and automatically generate supervision control strategy, to improve the real-time and intelligent level of engineering supervision.

Description

Technical Field

[0001] This invention relates to the field of engineering supervision technology, and in particular to a BIM-based intelligent management and control method and system for the entire engineering supervision process. Background Technology

[0002] Existing engineering supervision technologies mostly rely on manual inspections, video spot checks, or post-event verification based on single construction logs. They lack a unified dynamic analysis capability for the correlation between equipment operating status, actual construction behavior, and construction energy consumption at the construction site. Especially in complex engineering scenarios, traditional BIM supervision systems can usually only realize project progress display, component information management, and static process tracking. They are unable to perform fine-grained identification of the actual operating behavior of construction equipment and cannot accurately determine whether the construction process is consistent with the original engineering instructions.

[0003] Therefore, a BIM-based intelligent management and control method for the entire process of engineering supervision is needed. This method uses total energy consumption signals as a basis to identify the behavior of construction equipment and analyze the consistency of construction instructions, thereby achieving intelligent dynamic supervision and control throughout the entire construction process. Summary of the Invention

[0004] This invention aims to provide a BIM-based intelligent management and control method and system for the entire process of engineering supervision, which realizes reverse identification of the operating behavior of construction equipment based on total energy consumption signals, and further completes the consistency deviation analysis between construction behavior and engineering instructions, thereby realizing intelligent, dynamic and closed-loop engineering supervision and control for the entire construction process.

[0005] A BIM-based intelligent management and control method for the entire process of engineering supervision includes the following steps: Obtain the original engineering instruction data for the current construction phase; construct the planned construction behavior sequence and engineering behavior benchmark model based on BIM and the original engineering instruction data; predict the energy efficiency of the project based on the engineering behavior benchmark model to obtain the energy consumption benchmark fingerprint; The current construction phase involves several construction equipment. The total energy consumption signal is acquired. Based on the energy consumption benchmark fingerprint, the total energy consumption signal is constrained and decomposed to obtain the independent equipment operation characteristics corresponding to different construction equipment. Based on the independent equipment operation characteristics, the operation state sequence of each construction equipment is reconstructed, and the actual construction behavior sequence within the current construction phase is output. Consistency analysis is performed based on the actual construction behavior sequence and the planned construction behavior sequence to obtain the project consistency deviation result. Based on the project consistency deviation result, project supervision judgment is made, and the project supervision judgment result is obtained. Based on the assessment results of the engineering supervision, complete the engineering supervision and control of the current construction stage; based on the assessment results of the engineering supervision of all construction stages, complete the intelligent control of the entire engineering supervision process.

[0006] As a preferred technical solution of the present invention, the specific steps for constructing a planned construction behavior sequence and an engineering behavior benchmark model based on BIM and original engineering instruction data include: Semantic extraction is performed on the original engineering instruction data to obtain the original engineering action predicate, the original engineering action object noun, and the original engineering action time quantifier; engineering component entities are obtained based on BIM; the original engineering action object noun and the engineering component entities are semantically aligned to obtain the engineering component entity identifier code; The original engineering action predicates, original engineering action time quantifiers, and engineering component entity identifiers are used to perform process reasoning in a pre-constructed engineering supervision knowledge graph to obtain the corresponding planned construction behavior sequence; the engineering construction equipment combination and engineering period quota required for the corresponding planned construction behavior sequence under standard conditions are extracted. A benchmark model of engineering behavior is constructed based on the entity identification code of engineering components and the corresponding planned construction behavior sequence.

[0007] As a preferred embodiment of the present invention, the specific steps for predicting engineering energy efficiency based on an engineering behavior benchmark model include: Based on the entity identification code of engineering components, engineering environmental constraint parameters related to the corresponding planned construction behavior sequence are extracted from the engineering behavior benchmark model; an engineering power efficiency model is established based on the combination of engineering construction equipment; the engineering schedule quota is nonlinearly corrected based on the engineering environmental constraint parameters to obtain the corrected engineering schedule quota; the expected load rate and power fluctuation range of each engineering construction equipment in the current construction stage are calculated based on the corrected engineering schedule quota. Based on the expected load rate and power fluctuation range of each construction equipment in the current construction phase and the revised project schedule quota, power evolution curves of each type of equipment in the time domain are constructed; based on the power evolution curves, fast Fourier transform is performed to obtain frequency domain feature components; the power evolution curves and frequency domain feature components are combined and encapsulated to obtain the energy consumption benchmark fingerprint.

[0008] As a preferred embodiment of the present invention, the specific steps for constraining the total energy consumption signal of an engineering project based on an energy consumption benchmark fingerprint include: The total energy consumption signal of the project is sampled at high frequency to obtain a joint time-frequency signal of project energy consumption that includes transient and steady-state characteristics of the project; an energy efficiency observation sequence based on the construction time axis is established based on the joint time-frequency signal of project energy consumption. Based on the combination of engineering construction equipment, the corresponding prior energy consumption benchmark template is extracted from the energy consumption benchmark fingerprint; the equipment participation constraint set is constructed according to the prior energy consumption benchmark template; the equipment participation constraint set is used as the equipment activation constraint, and the frequency domain feature components in the energy consumption benchmark fingerprint are used as feature matching constraints. Constraint decomposition of the time-frequency joint signal of engineering energy consumption is performed based on equipment activation constraints and feature matching constraints to obtain independent equipment operation characteristics that match the construction equipment of each project.

[0009] As a preferred embodiment of the present invention, the specific steps for reconstructing the operating state sequence of each engineering construction equipment based on the operating characteristics of independent equipment include: The energy consumption state space of an engineering project is constructed using the category of construction equipment as the first dimension, the construction time axis as the second dimension, and the operating characteristics of individual equipment as the third dimension. The energy consumption state space of an engineering project is probabilistically decoded to identify the action switching points of each construction equipment on the construction time axis. Based on the action switching points, the operating state modes of each construction equipment under different construction time slices are reconstructed. The construction time axis contains several construction time slices. Based on the combination of engineering construction equipment, the operating state modes of each engineering construction equipment under different construction time slices are fused by time correlation to obtain associated engineering construction equipment; when the operating state modes of several associated engineering construction equipment all meet the process triggering threshold within a preset overlapping time, the corresponding actual construction behavior event is synthesized; and all actual construction behavior events are combined in chronological order to obtain the actual construction behavior sequence. Obtain the scheduled start time, scheduled duration, and BIM spatial location information from the original engineering instruction data; calculate the optimal matching path between the actual construction behavior sequence and the planned construction behavior sequence on the nonlinear time axis based on the dynamic time warping algorithm to obtain the construction sequence alignment result; and combine the construction sequence alignment result from three dimensions—time deviation, strength deviation, and integrity deviation—to obtain the engineering consistency deviation result.

[0010] As a preferred embodiment of the present invention, the specific steps for making engineering supervision judgments based on engineering consistency deviation results include: Based on the engineering consistency deviation results, time deviation, construction intensity deviation, and construction integrity deviation are extracted; a comprehensive risk assessment is conducted based on the time deviation, construction intensity deviation, and construction integrity deviation to obtain the engineering supervision judgment result; and the engineering supervision judgment result triggers the supervision and control strategy for the next construction stage. The results of engineering consistency deviations, engineering supervision judgments, and supervision and control strategies at the current construction stage are written into the engineering supervision experience database. Based on the engineering supervision experience database, the supervision and control strategies for subsequent construction stages are optimized and matched to achieve intelligent control of the entire engineering supervision process.

[0011] A BIM-based intelligent management and control system for the entire process of engineering supervision, comprising: The engineering supervision prediction module includes a behavior modeling unit; the behavior modeling unit is used to acquire the original engineering instruction data of the current construction stage; based on BIM and the original engineering instruction data, a planned construction behavior sequence and an engineering behavior benchmark model are constructed; based on the engineering behavior benchmark model, the energy efficiency of the project is predicted to obtain the energy consumption benchmark fingerprint; The engineering supervision and analysis module includes an energy consumption analysis unit and an engineering control unit. The current construction phase involves several construction equipment. The energy consumption analysis unit acquires the total energy consumption signal of the project. Based on the energy consumption benchmark fingerprint, it performs constraint decomposition on the total energy consumption signal to obtain the independent equipment operation characteristics corresponding to different construction equipment. The engineering control unit reconstructs the operating state sequence of each construction equipment based on the independent equipment operation characteristics, outputting the actual construction behavior sequence within the current construction phase. Consistency analysis is performed based on the actual construction behavior sequence and the planned construction behavior sequence to obtain the engineering consistency deviation results. The engineering supervision decision-making module includes a decision-making unit and a strategy optimization unit. The decision-making unit is used to make engineering supervision judgments based on the engineering consistency deviation results and obtain the engineering supervision judgment results. The strategy optimization unit is used to complete the engineering supervision control of the current construction stage based on the engineering supervision judgment results. Based on the engineering supervision judgment results of all construction stages, the module completes the intelligent control of the entire engineering supervision process.

[0012] The present invention has the following advantages: 1. This invention constructs an intelligent management and control system for the entire process of engineering supervision based on BIM models, semantic parsing of engineering instructions, and engineering energy consumption data. Through the construction of planned construction behavior sequences, generation of energy consumption benchmark fingerprints, identification of equipment operation characteristics, reconstruction of actual construction behavior, and analysis of construction consistency deviations, it achieves automatic perception and intelligent supervision of the actual operation process on the construction site; it continuously acquires dynamic status information during the construction process, promptly detects construction deviations, and automatically generates supervision and control strategies, thereby improving the real-time performance and intelligence level of engineering supervision.

[0013] 2. This invention generates an energy consumption benchmark fingerprint using a planned construction behavior sequence and an engineering behavior benchmark model. It then constructs equipment activation constraints and feature matching constraints based on this fingerprint, decomposing the total energy consumption signal and extracting independent equipment operation features corresponding to each construction device from the total energy consumption data. This deeply integrates the BIM construction plan with the energy consumption analysis process, enabling the construction plan to directly participate in equipment identification and behavior reconstruction. This effectively reduces feature aliasing issues when multiple devices are operating simultaneously, improves equipment identification accuracy and construction behavior reconstruction precision, and provides a more reliable data foundation for subsequent construction deviation analysis and intelligent supervision judgment. Attached Figure Description

[0014] Figure 1This is a schematic diagram of a BIM-based intelligent management and control system for the entire process of engineering supervision, as used in an embodiment of the present invention. Detailed Implementation

[0015] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this invention.

[0016] Example 1: A BIM-based intelligent management and control method for the entire process of engineering supervision, comprising the following steps: The process involves: acquiring the original engineering instruction data for the current construction phase; constructing a planned construction behavior sequence and an engineering behavior benchmark model based on BIM and the original engineering instruction data; predicting engineering energy efficiency based on the engineering behavior benchmark model to obtain an energy consumption benchmark fingerprint; where Building Information Modeling (BIM) refers to a digital engineering model used to store information on building components, construction processes, spatial structures, and construction attributes; subsequently, extracting engineering instruction information corresponding to the current construction phase from systems related to the actual engineering plan, such as the construction management system, construction schedule system, and on-site equipment scheduling system, and summarizing them according to a unified data format to form the original engineering instruction data. The original engineering instruction data includes construction task names, construction procedure descriptions, construction area information, planned start time, planned end time, target construction components, construction equipment scheduling information, and construction responsibility entity information. Specific data items are determined based on the actual construction situation. During data acquisition, structured and unstructured text data from different systems are first read through data interfaces. Then, field mapping rules are used to uniformly encode data fields from different sources. Next, multi-source data is correlated and matched based on timestamps and construction area identifiers to eliminate duplicate records and abnormal null values. For unstructured construction text, word segmentation and syntactic parsing methods are used to extract construction actions, construction objects, and construction time descriptions. Word segmentation involves dividing continuous text into semantically independent word units, while syntactic parsing analyzes the semantic dependencies between words to identify the action entities and execution relationships in the construction instructions. After parsing, all data is converted into a unified engineering instruction sequence and stored according to construction stages, yielding the original engineering instruction data corresponding to the current construction stage. This provides basic data support for subsequent construction behavior identification and engineering supervision analysis.

[0017] The specific steps for constructing a planned construction behavior sequence and engineering behavior benchmark model based on BIM and original engineering instruction data include: Semantic extraction is performed on the original engineering instruction data to obtain the original engineering action predicate, the original engineering action object noun, and the original engineering action time quantifier; engineering component entities are obtained based on BIM; the original engineering action object noun and the engineering component entities are semantically aligned to obtain the engineering component entity identifier code; Semantic extraction processing is performed on the original engineering instruction data. Semantic extraction refers to identifying information with practical engineering significance, such as construction actions, construction objects, and construction time, from the engineering text. In the processing, the original engineering instruction data is first cleaned to remove duplicate symbols, invalid characters, and formatting tags. Then, a construction semantic recognition algorithm based on part-of-speech tagging is used to segment the text. Part-of-speech tagging assigns semantic attributes such as action words, object words, or time words to words in the text. Subsequently, based on the dependencies in the construction statements, the original engineering action predicates, original engineering action object nouns, and original engineering action time quantifiers are extracted. The original engineering action predicates represent construction behaviors, such as pouring, hoisting, or welding; the original engineering action object nouns represent the corresponding construction objects, such as beams, steel cages, or formwork structures; and the original engineering action time quantifiers represent the construction duration, construction time nodes, or construction sequence relationships. After semantic extraction, engineering component entities are obtained based on Building Information Modeling (BIM). These engineering component entities refer to actual engineering structural units with unique attribute identifiers in the BIM, including walls, floors, pipelines, supporting structures, and equipment foundations. Subsequently, the original engineering action object nouns and engineering component entities are semantically aligned. Specifically, the original engineering action object nouns are converted into standard engineering terms. Then, the name attributes, spatial attributes, and material attributes of the engineering component entities are extracted. Finally, a semantic vector similarity calculation method is used to match the two. This method measures semantic closeness by utilizing the distance relationship between words in a vector space. When the similarity exceeds a preset threshold, the corresponding engineering component entity is mapped to the target construction object, and the corresponding engineering component entity identifier code is output. The engineering component entity identifier code is a digital code in the BIM used to uniquely represent a component entity, uniquely characterizing the corresponding beam, slab, column, wall, or electromechanical component, etc. The original engineering action predicates, original engineering action time quantifiers, and engineering component entity identifiers are used to perform process reasoning in a pre-constructed engineering supervision knowledge graph to obtain the corresponding planned construction behavior sequence; the engineering construction equipment combination and engineering period quota required for the corresponding planned construction behavior sequence under standard conditions are extracted. The pre-constructed engineering supervision knowledge graph is an engineering relationship network built based on construction specifications, construction process standards, and engineering experience rules. It contains the relationships between construction actions, construction procedures, construction equipment, and construction conditions. During the identification process, the entity identifier of the engineering component is used as the retrieval entry point to locate the corresponding engineering component node in the knowledge graph. The original engineering action predicate is used to match the corresponding construction procedure node, and time constraints are established by combining the original engineering action time quantifiers. Then, a procedure reasoning method combining graph path search and rule-based reasoning is used to search for construction activities related to the current construction object layer by layer along the procedure dependency edges in the knowledge graph. The sequential constraints, spatial constraints, and resource constraints between procedures are verified according to preset procedure rules. For construction activities that meet the constraints, they are arranged and combined in chronological order to form a planned construction behavior sequence that conforms to construction logic, thereby realizing the automatic derivation process from engineering instructions to the construction plan behavior chain. It should be noted that the construction process of the pre-built engineering supervision knowledge graph is as follows: First, collect engineering construction specifications, construction organization design documents, supervision implementation details, equipment operation manuals, historical construction cases, and component attribute data from the building information model, and then perform a unified format conversion on the above data; Here, the knowledge graph refers to a structured relational network used to describe engineering entities and their relationships. It consists of entity nodes and relation edges. Entity nodes are used to represent objects such as construction procedures, construction equipment, engineering components, construction environmental conditions, and quality risk events, while relation edges are used to represent the logical relationships between different entities; Subsequently, natural language processing was performed on the original engineering text. First, a word segmentation algorithm was used to divide the construction statements into independent word units. Then, part-of-speech recognition was used to distinguish action words, object words, equipment words, and time words. Dependency parsing was used to identify the construction logic relationships between words. For example, in "tower crane hoisting precast beams," "hoisting" was identified as a construction action, "precast beams" as the construction object, and "tower crane" as the construction equipment. After that, the extracted construction terms were standardized, mapping different expressions to unified engineering terms. For example, "concrete pouring" and "concrete casting" were uniformly mapped to standard process words. After standardization, entity nodes were established in the knowledge graph, and the relationships between entities were constructed according to the construction logic, including "construction process corresponds to construction equipment," "construction process acts on engineering components," and "construction..." The system identifies relationships such as "preceding and subsequent dependencies between construction processes" and "the construction environment affecting construction processes." A graph-based storage method is then used to encode all entity nodes and relationships. Each entity node generates a unique identifier, and each relationship edge is accompanied by a relationship type and association weight. The association weight represents the importance of different relationships, specifically calculated based on the frequency of occurrence in historical construction cases, the strength of process associations, and the importance level of supervision rules. Finally, historical engineering data is used to iteratively correct the knowledge graph. Specifically, historical construction behaviors and actual supervision results are input into the graph, and the risk results and equipment matching corresponding to different construction behaviors are statistically analyzed. Based on the feedback results, the relationship weights and process association rules are dynamically adjusted, thus forming an engineering supervision knowledge graph that reflects the actual construction logic of engineering projects.

[0018] After obtaining the planned construction behavior sequence, each construction behavior in the sequence is first matched with the construction resource mapping relationship in the engineering supervision knowledge graph to determine the type of engineering construction equipment required to complete the corresponding construction behavior. Among them, the engineering construction equipment combination refers to the set of all equipment required to complete a certain construction behavior or construction procedure. For example, the concrete pouring procedure may correspond to concrete pumping equipment, vibration equipment, and transportation equipment, etc. Based on the engineering component type, construction scale, and construction conditions corresponding to the construction behavior, the corresponding standard construction parameters are retrieved from the pre-established construction quota database. The construction quota database records the equipment configuration rules, unit project construction time, and resource consumption parameters for different procedures under standard construction conditions. Based on the construction sequence in the planned construction activities, the equipment configuration results corresponding to each construction activity are merged to eliminate duplicate equipment and establish equipment coordination relationships, forming the engineering construction equipment combination required for the current construction stage. Simultaneously, based on the unit project construction time, project scale, and process connection relationships corresponding to each construction activity, the standard operating time for each construction activity is calculated, and a construction time network is established by combining the constraints of allowed parallel execution and mandatory sequential execution between processes. Finally, critical path analysis is used to calculate the key process links affecting the overall construction cycle, and the cumulative construction time on the key process links is statistically analyzed to obtain the corresponding project duration quota, thereby providing a standardized construction basis for subsequent project energy efficiency prediction and energy consumption benchmark fingerprint construction.

[0019] The pre-established construction quota database primarily draws on industry-published engineering quota standards, historical project data from construction companies, and technical parameter data for construction equipment. Engineering quota standards can be derived from building construction budget quotas, installation engineering budget quotas, municipal engineering quotas, and internal company construction quotas. This data provides standard working hours, standard resource consumption, and standard equipment configuration rules for different construction procedures. Historical project data from completed projects includes construction logs, progress records, equipment operation records, and project acceptance documents, used to obtain the execution time of procedures and equipment usage during actual construction. Technical parameter data for construction equipment comes from equipment manufacturers' manuals, rated power parameters, operating efficiency parameters, and equipment operating specifications, used to establish a mapping relationship between equipment capabilities and construction tasks. Through the fusion of data from multiple sources, the construction quota database can both provide a basis for industry standards and reflect the characteristics of actual construction.

[0020] The construction of the construction quota database includes four stages: data collection, data standardization, process mapping, and quota generation. First, it collects construction process names, construction object types, construction quantities, construction durations, equipment usage records, and resource consumption records from historical engineering projects, and converts them into a standardized data format. Then, it establishes a construction process coding system, mapping construction activities with different expressions but the same meaning across different projects to a unified process category; for example, "beam rebar tying" and "rebar tying operation" are uniformly mapped to the rebar tying process. Next, it establishes the correspondence between component types and construction processes based on the BIM component classification system, and establishes the resource mapping relationship between construction processes and construction equipment. Based on this, it uses a large number of construction records of similar processes from historical projects as samples to statistically analyze construction time, equipment configuration quantity, and resource consumption, and calculates standard values ​​after removing abnormal data. The quartile interval method can be used to identify abnormal samples and delete data records that exceed the normal distribution range. Finally, a weighted average algorithm is used to calculate the standard construction time and standard equipment configuration parameters for different processes, with weights determined based on project scale, construction environment similarity, and construction quality level.

[0021] An engineering behavior benchmark model is constructed based on the entity identification code of the engineering component and the corresponding planned construction behavior sequence. The engineering behavior benchmark model is a digital model used to characterize the standard construction behavior characteristics of a specific engineering component under a specific construction procedure. In essence, it is a correlation mapping model between engineering components, construction procedures, construction equipment, construction resources, and construction sequence relationships.

[0022] First, the corresponding engineering component is located from the BIM model using the entity identification code of the engineering component, and the structural attribute parameters, geometric dimension parameters, spatial location parameters, and material attribute parameters of the engineering component are extracted. Among them, the structural attribute parameters are used to characterize the component category and stress characteristics, the geometric dimension parameters are used to characterize the engineering quantity information such as the component volume, area or length, the spatial location parameters are used to characterize the floor, area and spatial adjacency relationship of the component, and the material attribute parameters are used to characterize the construction object characteristics such as concrete grade and steel specifications.

[0023] A mapping relationship is established between each construction behavior node in the planned construction behavior sequence and its corresponding engineering component, and a directed network of construction behaviors is constructed according to the order of occurrence. In this network, each construction behavior node is associated with a corresponding combination of construction equipment, standard construction period, and resource consumption parameters. A multi-layered relationship of "engineering component—construction behavior—construction equipment—construction period" is established based on the construction quota database, and the importance of different construction behaviors is quantified using a behavior weight calculation method. The behavior weight can be comprehensively calculated based on the proportion of construction period, resource consumption, and the criticality of the project. Finally, the characteristics of engineering components, construction behaviors, equipment combinations, and time constraints are uniformly encoded to form an engineering behavior benchmark model that reflects the construction execution law under standard construction conditions, providing behavioral basis data for subsequent engineering energy efficiency prediction.

[0024] The specific steps for predicting engineering energy efficiency based on engineering behavior benchmark models include: Based on the entity identification code of engineering components, engineering environmental constraint parameters related to the corresponding planned construction behavior sequence are extracted from the engineering behavior benchmark model; an engineering power efficiency model is established based on the combination of engineering construction equipment; the engineering schedule quota is nonlinearly corrected based on the engineering environmental constraint parameters to obtain the corrected engineering schedule quota; the expected load rate and power fluctuation range of each engineering construction equipment in the current construction stage are calculated based on the corrected engineering schedule quota. Among them, engineering environmental constraint parameters refer to the set of external environmental factors that affect the operating efficiency of construction equipment, construction cycle, and resource consumption level, and are used to reflect the differences between actual construction conditions and standard construction conditions. First, the corresponding engineering components and their associated construction behavior links are located in the engineering behavior benchmark model using the entity identification code of the engineering components. Then, the construction environment information related to the construction behavior link is extracted. The construction environment information mainly includes construction space constraint parameters, construction area access constraint parameters, material transportation distance parameters, component installation height parameters, and construction area operation density parameters. Among them, construction space constraint parameters are used to reflect whether the equipment operating space is restricted; construction area access constraint parameters are used to reflect the difficulty of moving construction vehicles and construction equipment; material transportation distance parameters are used to characterize the average distance of transporting construction materials from the storage area to the construction area; component installation height parameters are used to reflect the efficiency changes brought about by high-altitude construction; and construction area operation density parameters are used to reflect the concentration of construction resources in the same area. Subsequently, the above parameters are normalized to convert data of different dimensions to a unified evaluation interval, and an environmental impact coefficient mapping table is established based on historical construction samples. By finding the mapping relationship, the degree of influence of each environmental parameter on construction efficiency is calculated, and finally, a set of engineering environmental constraint parameters is formed to correct the construction period and equipment load level.

[0025] The engineering dynamic energy efficiency model is a mathematical model used to describe the relationship between energy consumption and construction output of construction equipment during the execution of construction tasks. Its function is to establish the correspondence between construction behavior and equipment energy consumption. First, all equipment types, rated power parameters, rated operating efficiency parameters, and load characteristic parameters of the engineering construction equipment combination are obtained. Among them, the rated power parameter is used to characterize the power level of the equipment when it is running at full load, the rated operating efficiency parameter is used to characterize the ability to complete construction tasks per unit time, and the load characteristic parameter is used to characterize the relationship between equipment load changes and power changes. Based on the collaborative operation relationship between construction equipment, an equipment correlation matrix is ​​established to describe whether there are dependency relationships or synchronous operation relationships among the equipment during the construction process. The theoretical working time of each piece of equipment is calculated based on the equipment's workload and operating efficiency, and the operating sequence and overlapping operating time of each piece of equipment within the construction cycle are determined by combining the equipment correlation matrix. Further, a mapping function between equipment load rate and equipment power is established. By inputting the equipment load rate into the mapping function, the corresponding operating power is calculated, and finally, an equipment power output model covering the entire construction cycle is obtained. By superimposing and coupling analysis of all equipment power output models, an engineering dynamic energy efficiency model that can reflect the energy consumption change law of the entire construction process is formed.

[0026] The engineering construction period quota refers to the theoretical construction time required to complete a specific construction task under standard construction conditions. However, in actual construction, environmental factors often affect the construction period, causing deviations that require correction. First, the environmental impact factors in the engineering environmental constraints are obtained, and their corresponding environmental impact coefficients are calculated. The impact of different environmental factors on construction efficiency is not linear; for example, increasing construction height does not necessarily lead to a fixed increase in construction time. Therefore, a nonlinear correction method is used. Specifically, an environmental impact model can be established, using parameters such as construction space constraints, transportation distance, installation height, and work density as input variables, and using actual construction period deviations from historical engineering samples as training targets. The environmental impact model is constructed through nonlinear regression analysis. The engineering environmental constraints corresponding to the current construction stage are then input into the environmental impact model to calculate the comprehensive environmental correction coefficient. Finally, the comprehensive environmental correction coefficient is used to correct the standard engineering construction period quota, resulting in a corrected engineering construction period quota that better reflects the current construction site conditions, thereby improving the consistency between subsequent energy efficiency predictions and actual construction conditions.

[0027] In this embodiment, the correction of the engineering period quota by engineering environmental constraint parameters is based on the objective engineering law that the construction environment affects construction efficiency. Specifically, firstly, a large amount of environmental data, such as construction space conditions, material transportation conditions, component installation conditions, and resource distribution in the construction area, is collected from completed projects, and the actual construction period data of the corresponding processes are obtained simultaneously. Then, the environmental data is standardized to establish a correlation sample set between environmental factors and actual construction period deviations. Based on the correlation sample set, a nonlinear regression analysis method is used to explore the influence of each environmental factor on construction efficiency and construct an environmental impact model. Then, the engineering environmental constraint parameters corresponding to the current construction stage are input into the environmental impact model to calculate the comprehensive impact of environmental conditions on construction efficiency, and the standard engineering period quota is corrected accordingly to obtain a corrected engineering period quota that matches the current construction environment. This method can fully consider the dynamic changes in construction efficiency under different construction environments, making the obtained corrected engineering period quota more consistent with the actual construction situation, thereby improving the accuracy and reliability of subsequent equipment load prediction, energy consumption prediction, and energy consumption benchmark fingerprint construction results.

[0028] The expected load factor is the ratio between the expected workload of construction equipment during the current construction phase and its rated working capacity, used to characterize the degree of equipment utilization; the power fluctuation range is the range of power changes that may occur during the construction process, used to reflect the characteristics of changes in the equipment's operating status.

[0029] First, the actual execution time window for each construction task is determined based on the revised project schedule quota. Then, the task allocation for each piece of equipment is calculated by combining the equipment capacity parameters in the project construction equipment combination. The expected load rate of the equipment is calculated based on the ratio between the task allocation and the rated operating capacity of the equipment. A correspondence model between load rate and power output is established based on the historical operating data of the equipment. The operating power of the equipment under different load levels is calculated using the correspondence model, and the power change characteristics of the equipment in the start-up phase, stable operation phase, and load adjustment phase are statistically analyzed. Based on the power distribution results of each operating phase, the upper and lower boundary values ​​of the equipment power are calculated to form the power fluctuation range of the corresponding equipment in the current construction phase. This provides basic data support for the subsequent construction of power evolution curves and the generation of energy consumption benchmark fingerprints.

[0030] Based on the expected load rate and power fluctuation range of each construction equipment in the current construction phase and the revised project schedule quota, a power evolution curve for each type of equipment in the time domain is constructed; a fast Fourier transform is performed on the power evolution curve to obtain the frequency domain feature components; the power evolution curve and the frequency domain feature components are combined and encapsulated to obtain the energy consumption benchmark fingerprint. A power evolution curve is a time series curve used to describe the power variation of construction equipment throughout the entire construction cycle. It reflects the dynamic changes in energy consumption from startup, operation to shutdown. The process involves obtaining the expected load rate, power fluctuation range, and revised project schedule quota for each piece of construction equipment, and determining the expected operating time range of the equipment in the current construction phase based on the revised project schedule quota. The entire construction phase is then divided into several consecutive construction time slices according to the order of construction tasks, and the corresponding load level is calculated based on the amount of construction tasks undertaken by the equipment in different construction time slices. Next, a mapping relationship between load rate and actual power is established using historical equipment operating data, and the real-time power value corresponding to the equipment is calculated based on the load level in each time slice. For equipment with collaborative construction relationships, the power change process needs to be adjusted according to the collaborative operation rules of the equipment combination to reflect the linkage effect between equipment. Finally, the power values ​​corresponding to each time slice are continuously connected in chronological order to form a power evolution curve that characterizes the dynamic changes in equipment energy consumption. This curve not only preserves the overall energy consumption trend during equipment operation but also records the local power fluctuation characteristics caused by changes in equipment load, process switching, and changes in construction rhythm.

[0031] The Fast Fourier Transform (FFT) is an analytical method that converts time-domain signals into frequency-domain signals. Its function is to extract hidden periodic and frequency features from complex power change curves. Since the power evolution curve is a continuous power sequence arranged in time order, the power evolution curve is first sampled at equal time intervals to form a discrete power data sequence. Then, the discrete power data sequence is input into the Fast Fourier Transform algorithm, which calculates the contribution of different frequency components to the overall power signal by decomposing it layer by layer, thereby realizing the conversion of time-domain information to frequency-domain information.

[0032] During the transformation process, the algorithm first calculates the low-frequency components in the power signal to reflect the overall load change trend formed during long-term equipment operation; then it extracts the mid-frequency components to reflect the periodic characteristics formed by process switching and changes in construction rhythm; and further extracts the high-frequency components to reflect transient behavioral characteristics such as equipment startup, shutdown, and load surges. Afterwards, it statistically analyzes the energy distribution, dominant frequency position, and frequency amplitude variation patterns corresponding to each frequency component, forming frequency domain feature components that characterize the equipment's operating characteristics. Since different construction equipment has different operating modes and load change patterns, its corresponding frequency domain feature components usually have good discriminative ability and can serve as an important basis for identifying the equipment's operating status.

[0033] Energy consumption benchmark fingerprints are a set of digital features used to characterize the standard energy consumption characteristics of specific construction equipment or specific construction behaviors, functioning similarly to an energy consumption identifier during equipment operation. First, time-domain feature parameters reflecting the equipment's operating patterns are extracted from the power evolution curve, including average power, peak power, power fluctuation amplitude, operating duration, and power change trends. Then, dominant frequency features, frequency energy distribution features, frequency change patterns, and frequency stability features are extracted from the frequency-domain feature components. To eliminate the impact of differences in power levels between different equipment, the aforementioned time-domain and frequency-domain features need to be uniformly normalized. The data is converted to the same data scale range; then, the time domain features and frequency domain features are fused and encoded according to the preset feature encoding rules, and the correlation between corresponding equipment types, construction process types and construction environmental conditions is established; finally, the fused feature parameter set, equipment type identification information, construction behavior identification information and environmental constraint information are jointly encapsulated to form an energy consumption benchmark fingerprint; this energy consumption benchmark fingerprint not only retains the time domain change pattern and frequency domain change pattern during equipment operation, but also reflects the impact of construction process and construction environment on energy consumption characteristics, thus providing a reliable prior reference for the constraint decomposition of the total energy consumption signal of the subsequent project and the identification of the operating characteristics of construction equipment.

[0034] The current construction phase involves several engineering construction equipment; the total energy consumption signal of the project is obtained; the total energy consumption signal of the project is constrained and decomposed based on the energy consumption benchmark fingerprint to obtain the independent equipment operation characteristics corresponding to different engineering construction equipment; The specific steps for constraining the total energy consumption signal of the project based on the energy consumption benchmark fingerprint include: The total energy consumption signal of the project is sampled at high frequency to obtain a joint time-frequency signal of project energy consumption that includes transient and steady-state characteristics of the project; an energy efficiency observation sequence based on the construction time axis is established based on the joint time-frequency signal of project energy consumption. Based on the combination of engineering construction equipment, the corresponding prior energy consumption benchmark template is extracted from the energy consumption benchmark fingerprint; the equipment participation constraint set is constructed according to the prior energy consumption benchmark template; the equipment participation constraint set is used as the equipment activation constraint, and the frequency domain feature components in the energy consumption benchmark fingerprint are used as feature matching constraints. Constraint decomposition of the time-frequency joint signal of engineering energy consumption is performed based on equipment activation constraints and feature matching constraints to obtain independent equipment operation characteristics that match the construction equipment of each project.

[0035] Construction equipment refers to mechanical or power equipment used to perform specific construction procedures, including tower cranes, concrete conveying equipment, steel bar processing equipment, welding equipment, excavating equipment, and power generation equipment. During the identification process, the planned construction behavior sequence information corresponding to the current construction stage is first read, and then the corresponding construction equipment combination is matched according to the planned construction behavior sequence. Combining on-site equipment location data, equipment start-up and shutdown status data, and equipment operating area information, construction equipment in actual operation is screened. Equipment location data comes from equipment positioning modules or on-site wireless sensing terminals, and equipment start-up and shutdown status data comes from equipment controllers or energy consumption monitoring terminals. Then, the construction task time is aligned with the equipment operating time using timestamp synchronization to identify the set of equipment actually participating in construction during the current construction stage. Finally, an association table of construction equipment for the current construction stage is established based on equipment type, equipment number, equipment power level, and the construction area to which the equipment belongs, thereby determining the number of construction equipment included in the current construction stage.

[0036] During the current construction phase, multiple construction equipment typically operate simultaneously at the construction site. Smart meters, power distribution monitoring terminals, and fuel flow monitoring devices deployed at the site collect various energy data in real time. The total energy consumption signal reflects the overall energy consumption status of the entire construction area within the current time period. Specifically, it generally includes both electrical and fuel consumption. Electrical energy consumption reflects the real-time power demand changes of electrical equipment, while fuel consumption reflects the fuel consumption changes of fuel-powered equipment during operation. The collected raw energy consumption data undergoes time synchronization processing, mapping data from different acquisition devices to the same construction timeline. Anomaly detection and noise filtering are then performed to remove abnormal data caused by communication fluctuations, equipment malfunctions, or momentary interference, forming a continuous energy consumption monitoring sequence at uniform time intervals. Since most equipment at the construction site does not have independent energy metering devices, the final result is a total energy consumption data aggregated from multiple devices, providing a foundational data source for identifying the operating status of each device from the total energy consumption.

[0037] The total energy consumption signal of the project is a mixed signal formed by the superposition of energy consumption from multiple construction equipment. It includes both steady-state energy consumption characteristics formed by long-term stable operation of the equipment and transient energy consumption characteristics generated during equipment start-up, shutdown, and load switching. In order to completely preserve this information, the total energy consumption signal of the project is first continuously collected using a high-frequency sampling method, so that subtle changes during equipment operation can be completely recorded. Then, the sampled energy consumption data is divided into time windows, and characteristic parameters such as power change rate, load fluctuation amplitude, and energy consumption change trend are extracted within each time window. Among them, the transient characteristics of the project are used to describe the short-term energy consumption mutation characteristics generated when the equipment state changes, and the steady-state characteristics of the project are used to describe the continuous energy consumption characteristics formed during the stable operation phase of the equipment.

[0038] By combining time-domain and frequency-domain analysis, energy consumption data in each time window are jointly processed to form a joint time-frequency signal of engineering energy consumption that simultaneously contains information on time and frequency changes. Furthermore, the time-frequency characteristics corresponding to all time windows are arranged and correlated according to the chronological order of the construction timeline to construct an energy efficiency observation sequence that reflects the energy consumption change pattern throughout the construction process, thereby providing a continuous observation basis for subsequent equipment identification.

[0039] The energy consumption baseline fingerprint is a set of standard energy consumption features generated during the planned construction phase based on the combination of construction equipment, construction technology, and construction environmental conditions. The prior energy consumption baseline template is a reference feature template directly related to the current construction phase, selected from the energy consumption baseline fingerprint. First, the combination of construction equipment that should participate in the current construction phase is determined based on the planned construction behavior sequence. Then, using equipment type, construction behavior type, and construction phase information as search conditions, the time-domain and frequency-domain features of the corresponding equipment are extracted from the generated energy consumption baseline fingerprint. Next, the energy consumption features corresponding to different equipment are classified and organized, and a mapping relationship between equipment type and energy consumption features is established. Furthermore, based on the construction procedures involved in the current construction phase, the set of equipment features that need to be analyzed is selected and combined to form a prior energy consumption baseline template that can characterize the equipment operation law under standard construction conditions, thus providing a priori reference for the subsequent energy consumption decomposition process.

[0040] The equipment participation constraint set is a set of constraint rules used to limit the scope of equipment that may participate in the operation during the current construction phase. Its function is to introduce construction plan information into the energy consumption analysis process; to establish a candidate equipment set based on the equipment type information recorded in the prior energy consumption benchmark template; and to determine whether each piece of equipment has the conditions to participate in construction during the current construction phase by combining the planned construction behavior sequence; to assign an active status mark to equipment that meets the requirements of the construction plan, and to assign an inactive status mark to equipment that should not participate in the current construction phase, thereby forming the equipment participation constraint set.

[0041] The set of equipment participation constraints is used as equipment activation constraints to limit the types of equipment allowed in the subsequent decomposition process. At the same time, the dominant frequency features, frequency energy distribution features, and frequency change patterns of the corresponding equipment are extracted from the energy consumption benchmark fingerprint to construct feature matching constraints. Since different construction equipment usually have relatively stable frequency features during operation, feature matching constraints can be used to determine whether the observed energy consumption signal meets the standard operating characteristics of a certain type of equipment, thereby improving the accuracy of equipment identification.

[0042] Constraint decomposition refers to the process of breaking down the total energy consumption signal formed by the superposition of multiple devices into multiple independent energy consumption signals of multiple devices, given some prior information. First, the joint time-frequency energy consumption signal of the project is input into the energy consumption analysis model as the object to be decomposed, and the device activation constraints and feature matching constraints are introduced into the decomposition process simultaneously. Then, the range of devices participating in the decomposition is limited according to the device activation constraints to avoid the misidentification of devices that do not belong to the current construction stage.

[0043] The total energy consumption signal is progressively decomposed using an iterative optimization method. In each iteration, the matching degree between the current decomposition result and the corresponding device's frequency domain characteristics is calculated, and the allocation ratio of each device's feature components is continuously adjusted based on the matching results. When the decomposition result and the standard frequency characteristics of the corresponding device reach a preset matching requirement, the corresponding device's characteristics are considered successfully identified. Multiple independent device operation feature sequences are extracted from the total energy consumption signal, each corresponding to a specific operational state change pattern of the construction equipment. Because the entire decomposition process is simultaneously constrained by both the construction plan and energy consumption characteristics, it effectively reduces the probability of misidentification in multi-device parallel construction environments and improves the accuracy and stability of independent device operation feature extraction.

[0044] Based on the independent equipment operation characteristics, the operation status sequence of each construction equipment is reconstructed, and the actual construction behavior sequence within the current construction stage is output. Consistency analysis is performed on the actual construction behavior sequence and the planned construction behavior sequence to obtain the project consistency deviation result. Based on the project consistency deviation result, the project supervision judgment is made to obtain the project supervision judgment result. The specific steps for reconstructing the operating status sequence of each construction equipment based on its independent operating characteristics include: The energy consumption state space of an engineering project is constructed using the category of construction equipment as the first dimension, the construction time axis as the second dimension, and the operating characteristics of individual equipment as the third dimension. The energy consumption state space of an engineering project is probabilistically decoded to identify the action switching points of each construction equipment on the construction time axis. Based on the action switching points, the operating state modes of each construction equipment under different construction time slices are reconstructed. The construction time axis contains several construction time slices. The engineering energy consumption state space is a multi-dimensional data representation structure used to uniformly describe the operational state characteristics of different construction equipment over different time periods. Its purpose is to organize the dispersed equipment operational characteristics into a set of states that reflects the overall construction process. The category of construction equipment serves as the first dimension, distinguishing different types of equipment, such as tower cranes, construction hoists, concrete pump trucks, rebar processing equipment, welding equipment, and excavating equipment. Because different equipment undertake different construction tasks, their operational patterns and energy consumption characteristics also differ significantly; therefore, equipment categories need to be organized as an independent dimension. For example, in a certain construction phase, tower cranes are mainly responsible for material lifting, while concrete pump trucks are mainly responsible for concrete pouring. Although they may operate simultaneously, their corresponding construction behaviors are completely different, thus requiring differentiation through the equipment category dimension.

[0045] The construction timeline serves as the second dimension, describing the changes in equipment operating status over time. Typically, the entire construction phase is divided into several consecutive construction time slices, each corresponding to a fixed time range; for example, it can be divided at the minute, ten-minute, or hour level. The construction timeline records the changes in equipment operating status within different time periods, thereby identifying behaviors such as equipment startup, shutdown, continuous operation, and process switching; for example, if a piece of equipment is in continuous operation in the morning and in standby mode in the afternoon, this change is reflected in the time dimension.

[0046] The independent equipment operating characteristics, as the third dimension, originate from the equipment-specific operating characteristics obtained after constraining the total energy consumption signal. These characteristics primarily describe the equipment's operating state itself. They can include information such as equipment power level, load factor, power change rate, frequency characteristics, operating duration, and energy consumption fluctuation patterns. For example, a tower crane may be in a high-load lifting state during one time period, while in an idle standby state during another. The independent equipment operating characteristics corresponding to these two states are significantly different.

[0047] In practice, the operational characteristics of individual construction equipment are first categorized according to their type. For example, lifting equipment, concrete construction equipment, transportation equipment, and rebar processing equipment are grouped into their respective categories. Then, the operational characteristics of each piece of equipment are mapped to their corresponding construction time points along a unified construction timeline, and the entire construction cycle is divided into several consecutive construction time slices. Next, within each construction time slice, the power level characteristics, power change rate characteristics, load change characteristics, and operating duration characteristics of the corresponding equipment are extracted and stored as dimensions of independent equipment operational characteristics. Finally, the equipment category dimension, construction time dimension, and operational characteristic dimension are used to jointly construct the project energy consumption state space, enabling the operational status of different equipment throughout the entire construction cycle to be expressed and analyzed within a unified data framework, thus providing a basic data structure for subsequent state identification.

[0048] Probabilistic decoding refers to the analytical process of identifying the location of equipment operating state changes from continuous energy consumption characteristics by utilizing the probability law of state change. In specific implementation, firstly, the equipment operating feature vector corresponding to each construction time slice is extracted from the engineering energy consumption state space, and an equipment state transition model is established based on historical equipment operating samples. The state transition model is used to describe the probability law of equipment changing from one operating state to another, such as the change law of equipment changing from standby state to start-up state, from start-up state to stable operating state, and from operating state to shutdown state. Then, the degree of change of operating characteristics between adjacent construction time slices is calculated, and the probability of occurrence of the corresponding operating state in each time slice is obtained in combination with the state transition model. Then, the maximum probability path search method is used to determine the state sequence that best matches the current operating characteristic change law from all possible state change paths. When the equipment state changes significantly, a state probability abrupt change phenomenon will appear at the corresponding time position, so the state abrupt change position can be identified as the action switching point. Through this process, key state change moments such as equipment start-up, shutdown, load switching, and process conversion can be automatically identified from continuous energy consumption data.

[0049] Based on the combination of engineering construction equipment, the operating state modes of each engineering construction equipment under different construction time slices are fused by time correlation to obtain associated engineering construction equipment; when the operating state modes of several associated engineering construction equipment all meet the process triggering threshold within a preset overlapping time, the corresponding actual construction behavior event is synthesized; and all actual construction behavior events are combined in chronological order to obtain the actual construction behavior sequence. Operational state modes are a state expression form used to describe the comprehensive operational state characteristics of equipment within a specific construction time slice. They not only reflect the current state of the equipment but also its operational intensity and trends. In practice, the equipment operation process is first segmented using identified action switching points, dividing it into multiple continuous state intervals. Then, within each state interval, operational indicators such as average power, load level, continuous operating time, and fluctuation amplitude are statistically analyzed. Next, the state intervals are classified according to preset state identification rules. For example, a continuous high-load operation state is identified as a construction operation mode, a continuous low-power operation state as a standby mode, a rapidly increasing power state as a startup mode, and a rapidly decreasing power state as a shutdown mode. Furthermore, the state category, state duration, and state intensity information are integrated to form the operational state mode for the corresponding construction time slice. This process transforms complex equipment energy consumption variation patterns into state expression results with clear engineering significance.

[0050] Temporal correlation fusion refers to the process of jointly analyzing the operating status of multiple devices by utilizing the collaborative construction relationships between devices. In specific implementation, the collaborative operation relationship of devices is first extracted from the engineering construction equipment combination corresponding to the planned construction behavior sequence, and a collaborative relationship network of devices is established. Then, the time intervals corresponding to the operating status modes of different devices are analyzed, and the length of common operation time and the degree of state synchronization between devices are calculated.

[0051] When multiple devices are simultaneously in the same construction operation mode within the same construction time slice, and the duration of their state meets the requirements of the preset construction procedure, these devices are considered to have a collaborative construction relationship. Then, the time overlap ratio, the degree of synchronous change of state, and the degree of correlation of operation intensity between the devices are further calculated, and the device correlation relationship is established based on the comprehensive correlation degree. Devices with stable collaborative relationships are merged to form a set of related engineering construction equipment, thereby reflecting the operation characteristics of multiple devices jointly participating in the same procedure in the actual construction process.

[0052] The process trigger threshold is a criterion used to determine whether a certain construction process has actually occurred. It is derived from the standard construction behavior characteristics recorded in the engineering behavior benchmark model. In specific implementation, the operating status modes of the construction equipment of the related project within the same construction time slice are first obtained, and the common operating time, average load level and state stability between the equipment are calculated. Then, the above results are matched and analyzed with the standard characteristics of the corresponding construction process in the engineering behavior benchmark model.

[0053] When the associated equipment maintains an operating state that meets the construction requirements within a preset time range, and the equipment combination relationship, operating intensity, and duration all meet the corresponding process triggering conditions, it is determined that the construction process has actually occurred. The actual construction behavior event is generated by combining the process type, construction object, and participating equipment information, and the corresponding occurrence time, end time, and construction area information are recorded, thereby realizing the transformation from equipment operating state to construction behavior event.

[0054] The actual construction behavior sequence is a temporal representation of the actual construction activities performed on the construction site. It reflects the order and execution of construction procedures. In practice, firstly, all actual construction behavior events are sorted according to the construction timeline, and a time chain of behavior events is established. Then, the process dependencies and temporal connections between adjacent behavior events are analyzed. Behavior events that belong to the same construction task and are executed consecutively are merged, while behavior events belonging to different construction objectives are recorded independently. Next, the behavior event chain is logically verified by combining the physical information of engineering components and construction area information to ensure that the order of construction behaviors conforms to the laws of engineering construction. Finally, all behavior events are connected in chronological order to form a complete actual construction behavior sequence, which is used to represent the actual construction process that occurs on the construction site.

[0055] Obtain the scheduled start time, scheduled duration, and BIM spatial location information from the original engineering instruction data; calculate the optimal matching path between the actual construction behavior sequence and the planned construction behavior sequence on the nonlinear time axis based on the dynamic time warping algorithm to obtain the construction sequence alignment result; and combine the construction sequence alignment result from three dimensions—time deviation, strength deviation, and integrity deviation—to obtain the engineering consistency deviation result.

[0056] In practice, the planned start time, planned end time, and planned duration of the construction tasks are first extracted from the original engineering instruction data. Then, the corresponding engineering components are located in the BIM model using their entity identifiers, and the floor, spatial coordinates, and boundary information of the construction area are extracted. The planned start time characterizes the scheduled start time of the construction task, the planned duration characterizes the planned construction period, and the BIM spatial location information characterizes the spatial area where the construction task should occur. Finally, the time and spatial information are mapped to the planned construction behavior sequence, providing a spatiotemporal reference benchmark for subsequent construction behavior alignment analysis.

[0057] Dynamic Time Warping (RTW) is a matching analysis method for comparing the similarity of two time series of different lengths or execution speeds. In its implementation, the actual construction behavior sequence and the planned construction behavior sequence are first converted into behavior feature sequences, and a behavior correspondence matrix is ​​established. Then, the degree of difference between any two behavior nodes in the two sequences is calculated, including differences in behavior type, occurrence time, and spatial location. Next, dynamic programming is used to progressively calculate the cumulative matching cost from the start to the end of the sequence, and the matching path with the minimum cumulative cost is found. This path can achieve the optimal correspondence between the two construction behavior sequences while allowing for time scaling and compression. The alignment result of the construction sequences is obtained based on the optimal matching path, thus solving the problem of difficulty in directly comparing the actual execution speed and the planned execution speed at the construction site.

[0058] The engineering consistency deviation results are used to comprehensively reflect the degree of deviation between the actual construction process and the planned construction process. In practice, firstly, the time deviation is calculated based on the construction sequence alignment results. This time deviation characterizes the differences between the actual start time, end time, and duration of construction and the planned construction time. Next, the intensity deviation is calculated, characterizing the differences between the actual equipment operating intensity, construction load level, and resource input level during construction and the planned state. Then, the integrity deviation is calculated, characterizing whether all planned construction activities were executed and whether there were any omitted or added procedures. Further, the deviation information in these three dimensions is standardized, and the importance of each dimension's deviation is calculated according to engineering supervision evaluation rules. Finally, the deviation information in each dimension is combined and analyzed to form an engineering consistency deviation result that comprehensively reflects the consistency of construction execution, providing a basis for subsequent engineering supervision judgments and control decisions.

[0059] Based on the assessment results of the engineering supervision, complete the engineering supervision and control of the current construction stage; based on the assessment results of the engineering supervision of all construction stages, complete the intelligent control of the entire engineering supervision process.

[0060] The specific steps for making engineering supervision judgments based on engineering consistency deviation results include: Based on the engineering consistency deviation results, time deviation, construction intensity deviation, and construction integrity deviation are extracted; a comprehensive risk assessment is conducted based on the time deviation, construction intensity deviation, and construction integrity deviation to obtain the engineering supervision judgment result; and the engineering supervision judgment result triggers the supervision and control strategy for the next construction stage. Comprehensive risk assessment is a process of quantitatively analyzing and determining the degree of deviation from the plan during construction. Its purpose is to identify whether there are risk factors affecting the quality, safety, or schedule of the project at the current construction stage. In practice, firstly, time deviation, construction intensity deviation, and construction integrity deviation are standardized to eliminate the impact of differences in the dimensions of different indicators; then, a mapping relationship between deviations and risk levels is established based on historical project supervision data, and corresponding evaluation weights are assigned according to the different degrees of impact of different deviation indicators on the project outcome.

[0061] For example, time deviations involving critical path processes typically have a higher impact, while deviations in general auxiliary processes have a relatively lower impact. Weighted fusion of various deviation indicators is performed to calculate a comprehensive risk assessment value, and the corresponding risk level is determined according to preset risk level classification rules. Furthermore, the source of risk and its potential impact are analyzed using an engineering supervision experience database to generate engineering supervision judgment results. These judgment results can include different categories such as normal construction status, minor deviation status, moderate risk status, and major risk status, reflecting the overall execution status of the current construction phase.

[0062] In this embodiment, the weights of each deviation indicator can be statistically determined based on historical engineering supervision data. Specifically, firstly, time deviations, construction intensity deviations, and construction integrity deviations from a large number of completed projects, along with corresponding engineering quality results, schedule results, and safety management results, are collected. Then, the correlation between each deviation indicator and the final project risk is analyzed, and the contribution level of each deviation indicator to the risk results is calculated. Higher weights are assigned to deviation indicators with a higher correlation to the risk results, and lower weights are assigned to deviation indicators with a lower correlation. Further, the final weight parameters can be determined using the analytic hierarchy process, entropy weighting, or historical sample training methods, with specific setting rules adaptively adjusted according to the actual project situation. The weights are dynamically updated according to a preset cycle, allowing them to be continuously optimized as engineering supervision experience accumulates. The weights obtained in this way can objectively reflect the actual impact of different deviation factors on project risk, improving the accuracy of the comprehensive risk assessment results.

[0063] Supervision and control strategies are a set of monitoring, inspection, and adjustment measures formulated to address risks during the construction process. Their purpose is to proactively intervene in potential problems and reduce subsequent construction risks. In practice, the risk level and type corresponding to the current construction stage are first determined based on the engineering supervision assessment results. Then, supervision measures matching the current risk level are retrieved from a pre-established engineering supervision experience database. For example, when there is a significant schedule deviation, the frequency of schedule inspections in the next stage can be increased; when there are abnormal construction intensity, the content of equipment operation status verification can be increased; when there are deviations in construction integrity, the acceptance management of key processes can be strengthened. Next, based on the planned construction behavior sequence, construction area information, and construction equipment combination information for the next construction stage, candidate supervision measures are adaptively screened, and corresponding supervision and control strategies are generated. Finally, the supervision and control strategies are distributed to the supervision management system or construction management system to achieve dynamic adjustments to the supervision focus, inspection content, and risk control measures for the next construction stage.

[0064] The results of engineering consistency deviations, engineering supervision judgments, and supervision and control strategies at the current construction stage are written into the engineering supervision experience database. Based on the engineering supervision experience database, the supervision and control strategies for subsequent construction stages are optimized and matched to achieve intelligent control of the entire engineering supervision process.

[0065] The engineering supervision experience database is a structured knowledge database used to store knowledge of the engineering supervision process and historical supervision cases. Its function is to accumulate supervision experience and support intelligent decision-making in subsequent construction stages. In specific implementation, the consistency deviation results of the current construction stage are first standardized and coded to form a unified deviation feature record. Then, the engineering supervision judgment results are converted into corresponding risk level and risk type identifiers. After that, the supervision and control strategies actually adopted in this stage are described in a structured manner, including information such as the content of supervision measures, implementation time, and implementation objects.

[0066] Further establish the correlation between deviation characteristics, supervision judgment results, and supervision control strategies, and classify and store them according to construction project type, construction procedure category, and construction environment characteristics. Ultimately, this will form an engineering supervision experience record containing the correlation between "deviation characteristics—risk judgment—supervision measures—implementation effect," and be written into the engineering supervision experience database to provide historical reference for subsequent supervision decisions.

[0067] Intelligent management and control of the entire construction process refers to a dynamic management mechanism that continuously optimizes supervision decisions by leveraging historical supervision experience throughout the construction process. In practice, it first acquires the planned construction behavior sequence, project environment information, and real-time supervision data for subsequent construction phases, forming a set of construction characteristics for the current phase. Then, it searches the engineering supervision experience database for historical supervision cases similar to the current construction characteristics, selecting several candidate experience cases by calculating the similarity of construction behavior characteristics, deviation characteristics, and risk characteristics. Next, it statistically analyzes the implementation effects of corresponding supervision and control strategies in the candidate cases and examines the impact of different supervision measures on risk control results. Furthermore, it prioritizes the candidate supervision strategies based on historical implementation effects and selects the optimal supervision and control strategy as the recommended solution for the current phase. As construction progresses, newly generated deviation results, supervision judgment results, and supervision measure implementation results are continuously added to the engineering supervision experience database, constantly expanding and improving it. Through this closed-loop mechanism of continuous accumulation, matching, and optimization, the supervision decision-making capability can be continuously enhanced as the project progresses, thus achieving intelligent supervision and control covering the entire construction process.

[0068] It should be noted that the engineering supervision experience database can be pre-built based on engineering supervision standards, corporate supervision systems, and historical engineering supervision cases. Specifically, firstly, construction risks are categorized into schedule risks, quality risks, safety risks, and resource allocation risks according to risk type; then, a risk level system is established based on the severity of risks, and risk levels are divided into multiple levels such as normal level, early warning level, general risk level, high risk level, and major risk level; then, corresponding supervision measures are configured for different risk types and risk levels. For example, for schedule early warning risks, measures such as increasing the frequency of construction progress inspections and strengthening the tracking of key processes are configured; for quality risks, measures such as special quality spot checks, acceptance of key processes, and review of construction techniques are configured; for safety risks, measures such as on-site safety inspections, verification of equipment operating status, and special inspections of hazardous operations are configured; and the effectiveness evaluation records of strategies are formed by combining the risk improvement effects after the implementation of each supervision measure in historical projects, thereby constructing the engineering supervision experience database.

[0069] When matching supervision measures is required, the corresponding risk type and risk level are first determined based on the current project supervision assessment results. Then, candidate supervision measures corresponding to the risk type and risk level are retrieved from the project supervision experience database. Next, considering the construction procedures, equipment combinations, environmental conditions, and implementation effects of similar historical cases in the current construction phase, the candidate supervision measures are adaptively screened and prioritized. Supervision measures with better risk control effects in similar project scenarios are prioritized as the current stage's supervision and control strategy. In this way, the selection of supervision measures not only matches according to preset rules but also optimizes based on historical supervision experience, making the supervision and control strategy more consistent with the actual project scenario and improving the pertinence and effectiveness of the entire process of supervision and control.

[0070] In the actual implementation of the solution, when different data sources have inconsistent units, a unified data standard system is first established to identify the dimensions and map the units of all raw data. For example, power data is uniformly converted to kilowatts, electrical energy data to kilowatt-hours, time data to minutes or hours, length data to meters, and fuel consumption to liters or kilograms. Then, data conversion is completed according to preset unit conversion relationships to form a unified dimension dataset. Next, a normalization method is used to eliminate the influence of differences in parameter numerical ranges. By calculating the relative position of each parameter in the sample space, data of different dimensions are mapped to a unified numerical range. Further, missing value compensation, outlier removal, and consistency verification are performed on the normalized data to ensure that the data quality meets the model calculation requirements. Finally, the standardized data is input into various analysis and prediction models for calculation. Through these three steps—unified unit conversion, dimension normalization, and data quality verification—data from different sources, formats, and units can be integrated and analyzed within the same computational framework, thereby improving the accuracy and stability of the intelligent control results throughout the entire engineering supervision process.

[0071] In this embodiment, the knowledge reasoning model in the engineering supervision knowledge graph can be constructed using a graph reasoning algorithm. Specifically, by collecting engineering construction specifications, construction organization design documents, construction process standards, and historical engineering cases, a network of relationships between engineering components, construction procedures, construction equipment, and construction resources is established. Subsequently, a relationship path search method is used to identify the standard construction process corresponding to different engineering components, and the reasoning path is determined based on the frequency of occurrence of the procedure execution order in historical projects. When the occurrence ratio of a certain reasoning path in historical samples reaches the preset credibility requirement, it is used as a valid procedure reasoning rule.

[0072] The behavioral weight parameters in the engineering behavior benchmark model can be obtained from historical engineering samples. The weight values ​​are determined by analyzing the impact of different construction behaviors on the project period, equipment energy consumption, and construction quality. Higher weights are assigned to construction behaviors with greater impact, and lower weights are assigned to construction behaviors with less impact. The weights are then adjusted using cross-validation of historical projects.

[0073] The environmental impact model in the engineering environmental constraint parameters can be constructed using historical engineering environmental data and actual construction period deviation data to build a nonlinear regression model, and the model parameters can be determined by minimizing the prediction error; the environmental sensitivity coefficient can be obtained through multiple rounds of training with historical samples, and the final parameter can be determined when the model prediction error tends to stabilize.

[0074] The deviation weights in the comprehensive risk assessment can be determined by combining the analytic hierarchy process with historical supervision cases. A judgment matrix is ​​established through expert scoring, and the weights are corrected by combining the statistical results of historical risk events. The risk level thresholds in the engineering supervision experience database can be determined based on the deviation distribution in historical engineering projects. Historical deviation samples are divided into normal intervals, warning intervals, risk intervals, and high-risk intervals according to statistical distribution, and the corresponding boundary points are used as risk level thresholds.

[0075] The process trigger threshold can be obtained by statistically analyzing the equipment running time, equipment load rate, and equipment coordination level corresponding to the actual occurrence of the process in historical construction processes, and the threshold range is determined by percentile statistics, thereby ensuring that the threshold has a basis in actual engineering.

[0076] Example 2: A BIM-based intelligent management and control system for the entire engineering supervision process (see [link]). Figure 1 As shown, it includes: The engineering supervision prediction module includes a behavior modeling unit; the behavior modeling unit is used to acquire the original engineering instruction data of the current construction stage; based on BIM and the original engineering instruction data, a planned construction behavior sequence and an engineering behavior benchmark model are constructed; based on the engineering behavior benchmark model, the energy efficiency of the project is predicted to obtain the energy consumption benchmark fingerprint; The engineering supervision and analysis module includes an energy consumption analysis unit and an engineering control unit. The current construction phase involves several construction equipment. The energy consumption analysis unit acquires the total energy consumption signal of the project. Based on the energy consumption benchmark fingerprint, it performs constraint decomposition on the total energy consumption signal to obtain the independent equipment operation characteristics corresponding to different construction equipment. The engineering control unit reconstructs the operating state sequence of each construction equipment based on the independent equipment operation characteristics, outputting the actual construction behavior sequence within the current construction phase. Consistency analysis is performed based on the actual construction behavior sequence and the planned construction behavior sequence to obtain the engineering consistency deviation results. The engineering supervision decision-making module includes a decision-making unit and a strategy optimization unit. The decision-making unit is used to make engineering supervision judgments based on the engineering consistency deviation results and obtain the engineering supervision judgment results. The strategy optimization unit is used to complete the engineering supervision control of the current construction stage based on the engineering supervision judgment results. Based on the engineering supervision judgment results of all construction stages, the module completes the intelligent control of the entire engineering supervision process.

[0077] It should be understood that those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims. Parts not described in detail in this specification are prior art known to those skilled in the art.

Claims

1. A BIM-based intelligent management and control method for the entire process of engineering supervision, characterized in that, Includes the following steps: Obtain the original engineering instruction data for the current construction phase; construct the planned construction behavior sequence and engineering behavior benchmark model based on BIM and the original engineering instruction data; predict the energy efficiency of the project based on the engineering behavior benchmark model to obtain the energy consumption benchmark fingerprint; The current construction phase involves several engineering construction equipment; the total energy consumption signal of the project is obtained; the total energy consumption signal of the project is constrained and decomposed based on the energy consumption benchmark fingerprint to obtain the independent equipment operation characteristics corresponding to different engineering construction equipment; Based on the independent equipment operation characteristics, the operation status sequence of each construction equipment is reconstructed, and the actual construction behavior sequence within the current construction stage is output. Consistency analysis is performed on the actual construction behavior sequence and the planned construction behavior sequence to obtain the project consistency deviation result. Based on the project consistency deviation result, the project supervision judgment is made to obtain the project supervision judgment result. Based on the assessment results of the engineering supervision, complete the engineering supervision and control of the current construction stage; based on the assessment results of the engineering supervision of all construction stages, complete the intelligent control of the entire engineering supervision process.

2. The BIM-based intelligent management and control method for the entire process of engineering supervision according to claim 1, characterized in that, The specific steps for constructing a planned construction behavior sequence and engineering behavior benchmark model based on BIM and original engineering instruction data include: Semantic extraction is performed on the original engineering instruction data to obtain the original engineering action predicate, the original engineering action object noun, and the original engineering action time quantifier; engineering component entities are obtained based on BIM; the original engineering action object noun and the engineering component entities are semantically aligned to obtain the engineering component entity identifier code; The original engineering action predicates, original engineering action time quantifiers, and engineering component entity identifiers are used to perform process reasoning in a pre-constructed engineering supervision knowledge graph to obtain the corresponding planned construction behavior sequence; the engineering construction equipment combination and engineering period quota required for the corresponding planned construction behavior sequence under standard conditions are extracted. A benchmark model of engineering behavior is constructed based on the entity identification code of engineering components and the corresponding planned construction behavior sequence.

3. The BIM-based intelligent management and control method for the entire process of engineering supervision according to claim 2, characterized in that, The specific steps for predicting engineering energy efficiency based on engineering behavior benchmark models include: Based on the entity identification code of engineering components, extract engineering environmental constraint parameters related to the corresponding planned construction behavior sequence from the engineering behavior benchmark model; establish an engineering power efficiency model based on the combination of engineering construction equipment; perform nonlinear correction on the engineering schedule quota based on the engineering environmental constraint parameters to obtain the corrected engineering schedule quota; calculate the expected load rate and power fluctuation range of each engineering construction equipment in the current construction stage based on the corrected engineering schedule quota. Based on the expected load rate and power fluctuation range of each construction equipment in the current construction phase and the revised project schedule quota, power evolution curves of each type of equipment in the time domain are constructed; based on the power evolution curves, fast Fourier transform is performed to obtain frequency domain feature components; the power evolution curves and frequency domain feature components are combined and encapsulated to obtain the energy consumption benchmark fingerprint.

4. The BIM-based intelligent management and control method for the entire process of engineering supervision according to claim 3, characterized in that, The specific steps for constraining the total energy consumption signal of the project based on the energy consumption benchmark fingerprint include: The total energy consumption signal of the project is sampled at high frequency to obtain a joint time-frequency signal of project energy consumption that includes transient and steady-state characteristics of the project; an energy efficiency observation sequence based on the construction time axis is established based on the joint time-frequency signal of project energy consumption. Based on the combination of engineering construction equipment, the corresponding prior energy consumption benchmark template is extracted from the energy consumption benchmark fingerprint; the equipment participation constraint set is constructed according to the prior energy consumption benchmark template; the equipment participation constraint set is used as the equipment activation constraint, and the frequency domain feature components in the energy consumption benchmark fingerprint are used as the feature matching constraint. Constraint decomposition of the time-frequency joint signal of engineering energy consumption is performed based on equipment activation constraints and feature matching constraints to obtain independent equipment operation characteristics that match the construction equipment of each project.

5. The BIM-based intelligent management and control method for the entire process of engineering supervision according to claim 4, characterized in that, The specific steps for reconstructing the operating status sequence of each construction equipment based on the independent equipment operating characteristics include: The energy consumption state space of an engineering project is constructed using the category of construction equipment as the first dimension, the construction time axis as the second dimension, and the operating characteristics of individual equipment as the third dimension. The energy consumption state space of an engineering project is probabilistically decoded to identify the action switching points of each construction equipment on the construction time axis. Based on the action switching points, the operating state modes of each construction equipment under different construction time slices are reconstructed. The construction time axis contains several construction time slices. Based on the combination of engineering construction equipment, the operating state modes of each engineering construction equipment under different construction time slices are fused by time correlation to obtain associated engineering construction equipment; when the operating state modes of several associated engineering construction equipment all meet the process triggering threshold within a preset overlapping time, the corresponding actual construction behavior event is synthesized; and all actual construction behavior events are combined in chronological order to obtain the actual construction behavior sequence. Obtain the scheduled start time, scheduled duration, and BIM spatial location information from the original engineering instruction data; calculate the optimal matching path between the actual construction behavior sequence and the planned construction behavior sequence on the nonlinear time axis based on the dynamic time warping algorithm to obtain the construction sequence alignment result; and combine the construction sequence alignment result from three dimensions—time deviation, strength deviation, and integrity deviation—to obtain the engineering consistency deviation result.

6. The BIM-based intelligent management and control method for the entire process of engineering supervision according to claim 5, characterized in that, The specific steps for making engineering supervision judgments based on engineering consistency deviation results include: Based on the engineering consistency deviation results, time deviation, construction intensity deviation, and construction integrity deviation are extracted; a comprehensive risk assessment is conducted based on the time deviation, construction intensity deviation, and construction integrity deviation to obtain the engineering supervision judgment result; and the engineering supervision judgment result triggers the supervision and control strategy for the next construction stage. The results of engineering consistency deviations, engineering supervision judgments, and supervision and control strategies at the current construction stage are written into the engineering supervision experience database. Based on the engineering supervision experience database, the supervision and control strategies for subsequent construction stages are optimized and matched to achieve intelligent control of the entire engineering supervision process.

7. A BIM-based intelligent management and control system for the entire process of engineering supervision, characterized in that, The system applies the BIM-based intelligent management and control method for the entire process of engineering supervision as described in any one of claims 1-6, including: The engineering supervision prediction module includes a behavior modeling unit; the behavior modeling unit is used to acquire the original engineering instruction data of the current construction stage; based on BIM and the original engineering instruction data, a planned construction behavior sequence and an engineering behavior benchmark model are constructed; based on the engineering behavior benchmark model, the engineering energy efficiency is predicted to obtain the energy consumption benchmark fingerprint; The engineering supervision and analysis module includes an energy consumption analysis unit and an engineering control unit. The current construction phase involves several construction equipment. The energy consumption analysis unit acquires the total energy consumption signal of the project. Based on the energy consumption benchmark fingerprint, it performs constraint decomposition on the total energy consumption signal to obtain the independent equipment operation characteristics corresponding to different construction equipment. The engineering control unit reconstructs the operating state sequence of each construction equipment based on the independent equipment operation characteristics, outputting the actual construction behavior sequence within the current construction phase. Consistency analysis is performed based on the actual construction behavior sequence and the planned construction behavior sequence to obtain the engineering consistency deviation results. The engineering supervision decision-making module includes a decision-making unit and a strategy optimization unit. The decision-making unit is used to make engineering supervision judgments based on the engineering consistency deviation results and obtain the engineering supervision judgment results. The strategy optimization unit is used to complete the engineering supervision control of the current construction stage based on the engineering supervision judgment results. Based on the engineering supervision judgment results of all construction stages, the module completes the intelligent control of the entire engineering supervision process.

Images