A project cost accounting management method and system for engineering supervision

Abstract

This invention belongs to the technical fields of engineering management and intelligent supervision, and discloses a project cost accounting management method and system for engineering supervision. The method includes inputting preprocessed data packets into edge computing nodes, using a noise filtering matrix algorithm for dynamic data cleaning, and outputting a structured data set with quality labels. Based on the structured data set with quality labels, the construction site is scanned using augmented reality glasses, and the actual coordinates are compared with the design coordinates of the building information model in real time. When the accumulated deviation exceeds the warning threshold, an unalterable acceptance evidence chain is generated. The incremental costs of multiple standardized visa document packages are accumulated, and time series analysis is performed using a long-short-term memory prediction model to output a project cost analysis report with risk levels. This method solves the problem of the lack of acceptance evidence chain in traditional cost accounting and overcomes the strong dependence of conventional prediction models on data distribution.

Description

Technical Field

[0001] The present invention belongs to the technical field of engineering management and intelligent supervision, and specifically relates to a project cost accounting management method and system for engineering supervision. Background Art

[0002] In modern engineering construction, project supervision, as a crucial management link in ensuring project quality, progress, and cost, has become a key indicator of improving comprehensive project management capabilities through its informatization and digitization. With the development of technologies such as building information modeling, augmented reality, edge computing, and artificial intelligence, an increasing number of engineering projects are incorporating digitally-based project management solutions to assist supervisors in achieving data-driven, scientific management under complex construction conditions. Regarding cost accounting management in project supervision, some existing systems already enable basic statistical and comparative analysis of costs at each stage by combining BIM models with construction log records, manual acceptance reports, and construction drawings. Furthermore, some research and practice are exploring the use of IoT terminals to collect field data and centrally process it on cloud computing platforms to support subsequent analysis and report generation of cost deviations.

[0003] Existing engineering supervision cost accounting management solutions still have room for improvement in responding to dynamic changes on the construction site, real-time data-driven deviation identification and cost analysis. Especially when multi-source data real-time collection and edge computing processing capabilities are insufficient, existing solutions usually rely on manual data entry or centralized post-processing modes, resulting in data lags and untimely information updates, making it difficult to promptly reflect the direct incremental impact of construction deviations on project costs. In addition, although existing systems can perform static comparisons based on BIM models and construction plans, their implementation of the dynamic integration of augmented reality with actual on-site construction data and the instant generation of a traceable acceptance evidence chain is relatively limited. As a result, subsequent links such as visa document generation, cost increment accounting, and risk level analysis rely more on experience judgment or manual adjustments, affecting the standardization and refinement of management. This is particularly evident in complex projects and distributed construction scenarios, and places higher demands on supervision units. Summary of the Invention

[0004] In view of the above problems, the present invention provides a project cost accounting management method and system for engineering supervision, which solves the problem of insufficient real-time processing of multi-source data at the construction site and difficulty in dynamically quantifying cost increments caused by deviations.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, the present invention provides a project cost accounting management method for engineering supervision, comprising the following steps:

[0007] S1. Collect raw data from engineering monitoring, perform signal strength screening and timestamp synchronization processing, and generate pre-processed data packets;

[0008] S2: Input the pre-processed data packet into the edge computing node, use the noise filtering matrix algorithm to perform dynamic data cleaning, and output a structured data set with quality labels;

[0009] S3. Based on a structured data set with quality labels, the construction site is scanned using augmented reality glasses, and the actual coordinates are compared with the design coordinates of the building information model in real time. When the accumulated deviation exceeds the warning threshold, an unalterable acceptance evidence chain is generated;

[0010] S4. Based on the deviation data in the acceptance evidence chain, the incremental cost is calculated using the compensation coefficient matrix and a standardized visa document package is automatically generated;

[0011] S5. Accumulate the incremental costs of multiple standardized visa document packages, perform time series analysis using a long-short-term memory prediction model, and output a project cost analysis report with risk levels.

[0012] Furthermore, in S1, the collected engineering monitoring raw data includes spatial positioning data, structural deformation data, material performance parameters and environmental status parameters.

[0013] Furthermore, the signal strength screening and timestamp synchronization processing are performed to generate a pre-processed data packet, including:

[0014] The signal strength and environmental noise level of each sensor node are extracted from the engineering monitoring raw data, and the screening score is calculated using a dynamic adaptive threshold model. The corresponding engineering monitoring raw data with a score below the dynamic threshold screening score is marked as invalid and discarded.

[0015] Use the reference clock node to compare with other node clocks, calculate the time offset, apply the weighted correction model to adjust the timestamps of all engineering monitoring raw data, apply the final integration formula to unify the format of engineering monitoring raw data, and output the preprocessed data packet.

[0016] Furthermore, in S2, a structured dataset with quality labels is output, including:

[0017] Perform quantum chromodynamics noise modeling on the preprocessed data packets, output quantum noise characteristic tensors, and construct turbulent fluid dynamics equations to solve the filtered data field;

[0018] By integrating the chaotic pulse synchronization mechanism with Lie group differential manifolds, the multi-source time domain alignment of the filtered data field is performed, and the fractional-order Kalman filter algorithm is used to eliminate the drift error of the filtered data field after time domain alignment to generate drift-suppressed data.

[0019] A dynamic weight network is constructed based on the spatial topological relationship of drift suppression data, and the consistency of the associated data sets between spatial topological associated nodes is checked to generate discrete quality grade labels, which are fused with the drift suppression data to output a structured data set with quality labels.

[0020] Furthermore, in S3, an unalterable acceptance evidence chain is generated, including:

[0021] A quantum random number generator is used to quantize and encode the three-dimensional coordinate vectors in a structured data set with quality labels to generate quantum entangled state data. The laser radar of augmented reality glasses is used to scan the construction site to capture point cloud data in real time.

[0022] Real-time temperature data from the construction site is collected and the cumulative deviation of the thermodynamic entropy constraint is calculated. When the cumulative deviation exceeds the warning threshold, the point cloud topology genus features of the point cloud data are extracted, a three-dimensional feature hash value is generated, and a spatial topology certificate is signed using a quantum key.

[0023] The three-dimensional feature hash value and spatial topology certificate are submitted to the alliance chain, and a verifiable delay function is performed on the public chain through the cross-chain oracle to generate zero-knowledge proof and build an unalterable acceptance evidence chain.

[0024] Furthermore, in S4, a standardized visa document package is automatically generated, including:

[0025] The deviation accumulation value and cloud topology defect characteristics are extracted from the acceptance evidence chain and encoded. After quantum signature verification, the deviation data is output and input into the quantum coprocessor for variational quantum feature solution to obtain the compensation eigenvalue.

[0026] Performing a Hadamard product operation on the compensation eigenvalue and the preset compensation coefficient tensor to obtain the incremental cost;

[0027] Call the distributed legal engine to load the standard engineering contract template library, take the incremental cost and deviation characteristics as input parameters, apply the Laplace operator constraints of the contract terms to perform least squares optimization solution, and output a standardized visa document package.

[0028] Furthermore, in S5, a project cost analysis report with risk levels is output, including:

[0029] Decode the incremental cost characteristics of multiple standardized visa document packages and output a multi-dimensional cost characteristic matrix;

[0030] The multi-dimensional cost feature matrix is ​​input into the molecular dynamics simulation engine for continuous homology analysis to generate a dynamic topological feature sequence;

[0031] The dynamic topological feature sequence is input into the long-short-term memory prediction model to predict the evolution in the time dimension, output the incremental cost risk time series prediction data, and inject quantum random perturbations to output the risk evolution path;

[0032] Perform spatial convergence analysis on the risk evolution path to obtain the dynamic state characteristic vector, and perform discrete risk level division. This is then integrated with the incremental cost trend to generate a project cost analysis report with risk levels.

[0033] In a second aspect, the present invention provides a project cost accounting management system for engineering supervision, comprising:

[0034] The data acquisition module is used to collect raw data from engineering monitoring, perform signal strength screening and timestamp synchronization processing, and generate pre-processed data packets;

[0035] The edge cleaning module is used to input pre-processed data packets into the edge computing node, perform dynamic data cleaning using a noise filtering matrix algorithm, and output a structured data set with quality labels;

[0036] The verification and evidence storage module, based on a structured data set with quality labels, uses augmented reality glasses to scan the construction site and compare the actual coordinates with the design coordinates of the building information model in real time. When the accumulated deviation value exceeds the warning threshold, an unalterable acceptance evidence chain is generated;

[0037] The visa generation module is used to automatically generate a standardized visa document package based on the deviation data in the acceptance evidence chain and calculate the incremental cost using the compensation coefficient matrix;

[0038] The risk prediction module is used to accumulate the incremental costs of multiple standardized visa document packages, perform time series analysis through long-short-term memory prediction models, and output a project cost analysis report with risk levels.

[0039] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of the project cost accounting management method for engineering supervision as described in the first aspect of the present invention is implemented.

[0040] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the project cost accounting management method for engineering supervision as described in the first aspect of the present invention.

[0041] The present invention achieves the following beneficial effects: Through a quantized deviation verification and evidence storage process, it enables dynamic detection of millimeter-level construction deviations and legally reliable evidence storage, resolving the issue of a missing chain of evidence for acceptance in traditional cost accounting. Furthermore, based on a dynamic risk prediction process, it transforms incremental cost data into an interpretable risk evolution path, overcoming the traditional prediction model's strong dependence on data distribution. These two collaborative approaches form a closed loop of "precision measurement-reliable evidence storage-intelligent decision-making," systematically overcoming the three major technical gaps in engineering cost accounting: measurement error transmission, lack of legal validity, and unexplainable risk. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A flowchart of the project cost accounting management method used for engineering supervision;

[0044] Figure 2 This is a schematic diagram of a project cost accounting management system used for project supervision;

[0045] Figure 3 This is the data preprocessing flow chart;

[0046] Figure 4 Generate a flow chart for the acceptance evidence chain. DETAILED DESCRIPTION

[0047] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0050] Reference Figures 1 to 4, is an embodiment of the present invention, which provides a project cost accounting management method for engineering supervision, comprising the following steps:

[0051] S1. Collect engineering monitoring raw data, perform signal strength screening and timestamp synchronization processing, and generate pre-processed data packets.

[0052] S1.1. The collected engineering monitoring raw data includes spatial positioning data, structural deformation data, material performance parameters and environmental status parameters.

[0053] Furthermore, the raw data collection process for engineering monitoring includes four specific data types: spatial positioning data, structural deformation data, material performance parameters, and environmental state parameters. Spatial positioning data is mainly obtained through the global navigation satellite system method, recording the latitude and longitude coordinates and altitude of the measured object, for example, by deploying positioning instruments at key points of the structure for real-time coordinate measurement. Structural deformation data is collected using deformation monitoring methods, reflecting the displacement or shape change of the structure under external loads. Material performance parameters are determined using material testing methods to characterize the physical and mechanical properties of the materials that constitute the structure. Environmental state parameters use environmental monitoring methods to record external influencing factors, including changes in conditions such as temperature, humidity, and wind speed.

[0054] S1.2. Extract the signal strength and environmental noise level of each sensor node from the engineering monitoring raw data, calculate the screening score through the dynamic adaptive threshold model, and mark the corresponding engineering monitoring raw data with a screening score below the dynamic threshold as invalid and eliminate it.

[0055] Furthermore, the signal strength and ambient noise level of each sensor node are extracted from the raw engineering monitoring data. Data parsing methods are used to directly read the specific values ​​of each sensor node from the raw engineering monitoring data. These values ​​are then fed into a dynamic adaptive threshold model, which uses statistical analysis to calculate an output screening score that reflects the signal reliability of the sensor node. The model then determines a dynamic threshold based on real-time signal characteristics. Data entries with screening scores below the dynamic threshold are marked as invalid. Data filtering methods are then applied to remove these invalid entries, retaining the remaining raw engineering monitoring data.

[0056] It should also be noted that the dynamic adaptive threshold model is trained using historical signal strength and noise data, using statistical methods to optimize parameters and establish a real-time signal reliability evaluation function to achieve adaptive screening. The real-time signal characteristics are directly derived from the real-time monitoring data of the sensor node's current and historical sliding windows of signal strength and ambient noise levels.

[0057] S1.3. Use the reference clock node to compare with other node clocks, calculate the time offset, apply the weighted correction model to adjust the timestamps of all engineering monitoring raw data, apply the final integration formula to unify the format of engineering monitoring raw data, and output the pre-processed data packet.

[0058] Furthermore, a reference clock node is used to compare with other node clocks. The reference time signal provided by the reference clock node and the time value recorded by each node clock are directly compared through the time synchronization method to calculate the time offset corresponding to each node. The network time protocol method is used to obtain the precise difference during the data acquisition process. The weighted correction model is applied to adjust the timestamps of all engineering monitoring raw data. The weighted correction model uses the time offset obtained based on the time synchronization method as input, assigns a weight factor to the clock data of each node, and performs weighted calculation to adjust the timestamp value in the engineering monitoring raw data. For example, the weight value is set according to the importance of different sensor nodes for adjustment. The final integration formula is applied to uniformly encapsulate the format of the engineering monitoring raw data. The final integration formula integrates the adjusted timestamps and the original data into a consistent data structure through standardized data processing methods, converts all engineering monitoring raw data into a unified timestamp-value pair format, and outputs a pre-processed data packet containing time-corrected and format-unified content.

[0059] It should also be explained that the training process of the weighted correction model uses the historical time offset data of the reference clock node and each sensor node as input, fits the clock drift characteristics of different nodes through the least squares method, establishes a weight distribution matrix based on the node importance classification, and uses ridge regression to optimize the timestamp correction coefficient. Finally, a weighted correction model is generated that can dynamically adjust the weight according to the node type and signal quality.

[0060] S2. Input the preprocessed data packet into the edge computing node, use the noise filtering matrix algorithm to perform dynamic data cleaning, and output a structured data set with quality labels.

[0061] S2.1. Perform quantum chromodynamics noise modeling on the preprocessed data packet, output the quantum noise characteristic tensor, and construct the turbulent fluid dynamics equation to solve the filtered data field.

[0062] Furthermore, the quantum chromodynamics noise modeling method is applied to process the preprocessed data packets. The quantum chromodynamics noise modeling calculates the noise distribution based on the quantum chromodynamics theory, extracts the noise parameter model by analyzing the time-space domain signal pattern, outputs the quantum noise characteristic tensor as the noise characteristic representation in the multidimensional array format, constructs the turbulent fluid dynamics equation, inputs the quantum noise characteristic tensor as the initial condition or boundary value, and the turbulent fluid dynamics equation adopts the standard Add turbulence terms to the equation form and use average The fluid force field model is set up by the equation, the filtered data field is solved, the turbulent fluid dynamics equation is analyzed by numerical solution method, the high-frequency noise component is removed by iterative calculation through the finite volume method, and the filtered data field is output.

[0063] S2.2. By integrating the chaotic pulse synchronization mechanism through Lie group differential manifolds, the multi-source time domain alignment of the filtered data field is performed, and the fractional-order Kalman filter algorithm is used to eliminate the drift error of the filtered data field after time domain alignment to generate drift-suppressed data.

[0064] Furthermore, a chaotic pulse synchronization mechanism is integrated into the Lie group differential manifold to perform multi-source time domain alignment. A chaotic pulse synchronization method is used to generate a synchronous pulse signal. After the synchronous pulse signal is generated, the time phase difference between different data sources in the filtered data field is extracted. The phase difference is then input into the Lie group differential manifold structure. The data sources are mapped to the same differential manifold space through group action transformation. The time coordinates are aligned along the geodesic path on the manifold. The phase difference is converted into a tangent space translation vector through Lie algebraic mapping to achieve time domain calibration.

[0065] It should also be explained that the fractional order Kalman filter algorithm is used to eliminate the drift error of the filtered data field after time domain alignment. The aligned filtered data field is used as the observation sequence to input the fractional order Kalman filter algorithm. The algorithm establishes the state equation and the observation equation through the fractional order integral operator, for example, The fractional-order derivatives are defined to construct the state transfer matrix, execute the prediction-correction iterative process of the fractional-order Kalman filter algorithm, update the state estimation covariance matrix and compensate for the drift accumulation error, and finally output the drift suppression data.

[0066] S2.3. Construct a dynamic weight network based on the spatial topological relationship of the drift suppression data, perform consistency check on the associated data sets between the spatial topological associated nodes, generate discrete quality grade labels, and fuse them with the drift suppression data to output a structured data set with quality labels.

[0067] Furthermore, a spatial topological relationship extraction method is applied to process drift suppression data, obtaining node positions and spatial distance matrices as input parameters. A dynamic weighted network is constructed based on the spatial topological relationship, and the connection weights between nodes are determined using a geometric distance calculation function. For example, when the distance between nodes is less than the dynamic adjacency determination threshold range (for example, 5 meters), a weight value of 1 is assigned, otherwise it is 0, forming a dynamic weighted network graph structure.

[0068] Perform consistency checks on the associated datasets between spatial topologically associated nodes. Input the associated node pairs defined by the dynamic weight network and use correlation analysis methods to compare data differences between the nodes. For example, calculate the covariance matrix to detect numerical deviations, output the consistency index, and generate discretized quality grade labels. Input the consistency index and apply the grading threshold method to set the discrete grade. For example, if the consistency index is higher than 0.8, it is marked as "high", if it is between 0.5 and 0.8, it is marked as "medium", and if it is lower than 0.5, it is marked as "low". Create a discretized quality grade label array.

[0069] Fuse the discretized quality grade labels with the drift suppression data. Attach the discretized quality grade labels as new data columns through a data field append operation. Integrate all information and output a structured dataset with quality labels in a standard table or array format.

[0070] It should also be explained that the spatial topological relationship is directly inherited from the spatial positioning data in the preprocessed data packet, and the relative positions between nodes are recalculated during the multi-source time domain alignment process performed by the chaotic pulse synchronization mechanism, and finally the explicit spatial connection relationship is generated through the Lie group differential manifold mapping; the associated dataset is essentially a node pairing monitoring dataset extracted from the drift suppression data according to the topological connection relationship of the dynamic weight network.

[0071] S3. Based on a structured data set with quality labels, the construction site is scanned through augmented reality glasses, and the actual coordinates are compared with the design coordinates of the building information model in real time. When the cumulative deviation value exceeds the warning threshold, an unalterable acceptance evidence chain is generated.

[0072] S3.1. Use a quantum random number generator to quantize the three-dimensional coordinate vectors in the structured data set with quality labels to generate quantum entangled state data, and use the lidar of augmented reality glasses to scan the construction site to capture point cloud data in real time.

[0073] Furthermore, the three-dimensional coordinate vectors in the structured data set with mass labels are extracted, a quantum random number generator is used to generate a random bit sequence, quantum gate operations are applied to encode the scalar values ​​of the three-dimensional coordinate components into a single quantum bit sequence, and quantum entanglement operations are performed to convert the single quantum bit sequence into multi-bit quantum entangled state data and store it as an amplitude vector.

[0074] The laser radar sensor of the augmented reality glasses emits pulsed lasers and receives reflected signals, calculates the three-dimensional coordinates of the spatial points of the actual construction site based on the flight time difference and the scanning angle, and directly outputs the point cloud data matrix of the construction site; finally, two independent outputs of quantum entangled state data and construction site point cloud data are generated synchronously.

[0075] It should also be explained that the three-dimensional coordinate vectors in the structured data set with quality labels are directly derived from the GNSS positioning measurement results in the original engineering monitoring data, and are merged with the quality labels generated by the dynamic weight network after timestamp correction.

[0076] S3.2. Real-time temperature data of the construction site is collected and the cumulative deviation of the thermodynamic entropy constraint is calculated. When the cumulative deviation exceeds the warning threshold, the point cloud topological defect features of the point cloud data are extracted, a three-dimensional feature hash value is generated, and a spatial topological certificate is signed using a quantum key.

[0077] Furthermore, a temperature sensor network was deployed at the construction site to acquire real-time temperature data at a sampling frequency of 1 Hz. The instantaneous entropy value was calculated based on the Boltzmann entropy formula, where the temperature data was binned into 1°C intervals for statistical probability distribution, and the entropy deviation was calculated. The expression is:

[0078] ;

[0079] in, express The entropy deviation at the moment, Indicates the maximum reference entropy value allowed by the design, represents the chronological entropy value at time t;

[0080] right The time integral is performed to obtain the entropy deviation. When the cumulative deviation exceeds the entropy deviation accumulation threshold, the topology authentication process is triggered: the Vietoris-Rips complex algorithm is used to extract the Betti numbers and Euler characteristic numbers from the construction site point cloud data to form a topological feature vector. A 256-bit feature summary is generated through a SHA-256 hash operation. Finally, a quantum key generated by the BB84 protocol is used for digital signature, and a spatial topology certificate containing a timestamp, entropy deviation data, a topological feature hash value, and a quantum signature is output.

[0081] It should also be explained that the entropy deviation accumulation threshold is calculated by calculating the maximum allowable entropy change value through the material specific heat capacity parameter and the allowable operating temperature range, and then multiplying it by the monitoring time to obtain a fixed threshold.

[0082] S3.3. Submit the three-dimensional feature hash value and spatial topology certificate to the consortium chain, and generate zero-knowledge proof on the public chain through a verifiable delay function using a cross-chain oracle to build an unalterable acceptance evidence chain.

[0083] Furthermore, the three-dimensional feature hash value and the spatial topology certificate are submitted to the alliance chain smart contract interface to generate the alliance chain certificate. , the cross-chain oracle monitors the certificate Event, execute the verifiable delay function on the public chain, and use the verifiable delay function (VDF) to check the input Execute Continuous The deterministic output generated after the square operation and its verification proof π, where is the final hash value result, A proof of computational correctness that can be quickly verified; and Generate a certificate for the input in the public chain zero-knowledge proof protocol and will keep the evidence 、 、 The sequence calculates the root hash value through the Merkle tree , and finally anchored to the designated block of the public chain to build an unalterable chain acceptance evidence chain.

[0084] S4. Based on the deviation data in the acceptance evidence chain, the incremental cost is calculated using the compensation coefficient matrix, and a standardized visa document package is automatically generated.

[0085] S4.1. Extract the cumulative deviation value and cloud topology defect characteristics from the acceptance evidence chain and encode them. After quantum signature verification, output the deviation data and input it into the quantum coprocessor to perform variational quantum feature solution to obtain the compensation eigenvalue.

[0086] Furthermore, the deviation accumulation value and point cloud topology defect features are extracted from the Merkle tree leaf nodes of the acceptance evidence chain, and then merged and The hash algorithm generates a 256-bit encoded digest and is executed using the quantum signature and public key in the spatial topology certificate. The protocol verification process outputs the deviation data after confirming that the quantum signature matches the encoded summary; the deviation data is input into the quantum coprocessor, and a Hamiltonian H is constructed with the deviation accumulation value as the coefficient and the point cloud topology genus feature as the basis function. The minimum expected value of H is iteratively solved through parameterized quantum circuits and classical optimizers, and finally the compensation eigenvalue calculated by the quantum coprocessor is output.

[0087] S4.2. Perform a Hadamard product operation on the compensation eigenvalue and the preset compensation coefficient tensor to obtain an incremental cost.

[0088] Furthermore, the compensation eigenvalues ​​output by the quantum coprocessor are expanded into a tensor of the same dimension as the preset compensation coefficient tensor, and the product of each position element is calculated through the Hadamard product operation: , and finally outputs an incremental cost tensor with exactly the same dimension as the compensation coefficient tensor;

[0089] It should also be explained that the preset compensation coefficient tensor is calculated by finite element simulation of the stress-strain distribution of the structural grid unit, and the mechanical parameters are mapped into spatial compensation coefficients in combination with the material constitutive relationship and engineering economic model and generated by encapsulation according to three-dimensional grid coordinates.

[0090] S4.3. Call the distributed legal engine to load the standard engineering contract template library, take the incremental cost and deviation characteristics as input parameters, apply the Laplace operator constraints of the contract terms to perform least squares optimization solution, and output the standardized visa document package.

[0091] Furthermore, the distributed legal engine is called to load the national standard engineering contract template library, and the incremental cost tensor is converted into nine groups of cost vectors according to spatial coordinate partitions. These are then combined with the deviation accumulation value and point cloud topological defect features in the acceptance evidence chain to form a deviation feature parameter group. Based on the contract template clause coefficient matrix and discrete Laplace operator constraints, the conjugate gradient method is used to solve the least squares optimization problem, outputting the nine-zone visa amount vectors and clause index numbers. Finally, the results are filled into the contract template, and the digital signature of the National Time Service Center is attached to encapsulate it into a standardized visa document package.

[0092] It should also be noted that the deviation feature is directly combined and encapsulated into a 17-parameter structure by the cumulative deviation value and the point cloud topology defect feature when generating the acceptance evidence chain, without any numerical transformation or calculation processing. The Laplace operator constraint of the contract terms is generated by loading the standard engineering contract template library in the distributed legal engine, extracting the clear amount gradient constraint clauses in the template as the gradient threshold of the compensation amount of adjacent partitions, and using the discrete Laplace operator in standard mathematics to quantify the spatial difference of the amount vectors of adjacent partitions to generate specific constraint conditions. ( is the discrete Laplace operator, c is the target vector of contract amount), and is used for least squares optimization solution.

[0093] S5. Accumulate the incremental costs of multiple standardized visa document packages, perform time series analysis using a long-short-term memory prediction model, and output a project cost analysis report with risk levels.

[0094] S5.1. Decode the incremental cost characteristics of multiple standardized visa document packages and output a multi-dimensional cost characteristic matrix.

[0095] Furthermore, starting from multiple standardized visa file packages, the visa amount vector field and clause index number field stored in each standardized visa file package are read, and the incremental cost feature decoding is performed on each standardized visa file package. The specific operation is through The parser deserializes the visa amount vector, extracts the amount value as the incremental cost feature, and assembles the incremental cost features of all standardized visa file packages into a multi-dimensional cost feature matrix in the order of file index;

[0096] It should also be stated that the implementation of multiple standardized visa document packages Parsing operation, extract the visa amount vector fields encapsulated in the file one by one, discard the clause index number field, and directly output the amount vector as the incremental cost feature.

[0097] S5.2. Input the multidimensional cost feature matrix into the molecular dynamics simulation engine for continuous coherence analysis to generate a dynamic topological feature sequence.

[0098] Furthermore, the multidimensional cost feature matrix is ​​input into the molecular dynamics simulation engine, and the matrix element values ​​are converted into molecular potential energy parameters through linear mapping rules. The spatial partition index is mapped into three-dimensional coordinates, and the setting Potential function force field, perform 100 picosecond molecular dynamics trajectory simulation; sample molecular configuration every 1 picosecond, apply The complex algorithm calculates the Betti number of the homology dimension at each time point, and counts the mean and variance of the barcode survival length; and outputs a dynamic topological feature sequence serialized by time.

[0099] S5.3. Input the dynamic topological feature sequence into the long-short-term memory prediction model to perform time dimension evolution prediction, output incremental cost risk time series prediction data, inject quantum random perturbations, and output the risk evolution path.

[0100] Furthermore, the dynamic topological feature sequence is input into the long-short-term memory prediction model to perform time dimension evolution prediction. The long-short-term memory prediction model calculates the cell state and hidden state through the forget gate, input gate, and output gate state update mechanism, outputs the incremental cost risk time series prediction data for the next 30 steps, connects to the quantum random number generator to generate 30 sets of true random numbers, and injects the random numbers proportionally into the incremental cost risk time series prediction data according to the perturbation formula, generating a dual-channel risk evolution path matrix containing the original prediction value and the perturbation value;

[0101] It should also be explained that the long-short-term memory prediction model uses 10,000 sets of dynamic topological feature sequences in the historical engineering monitoring database and the corresponding incremental cost deviation true values. It is iteratively trained for 100 rounds using the Adam optimizer with a mean square error loss function. During the training process, 20% of the validation set is reserved and the error requirement of ≤5% is met, and finally a trained long-short-term memory prediction model is generated.

[0102] S5.4. Perform spatial convergence analysis on the risk evolution path to obtain the dynamic state characteristic vector, and perform discrete risk level classification. This is integrated with the incremental cost trend to generate a project cost analysis report with risk levels.

[0103] Furthermore, a spatial convergence analysis is performed on the risk evolution path. The Lyapunov exponent calculation method is used to quantify the degree of trajectory divergence between adjacent time steps in the path matrix. The maximum Lyapunov exponent, attractor dimension, and steady-state probability distribution parameters are extracted to form the dynamic state eigenvector. Based on the K-means clustering algorithm, the dynamic state eigenvector is divided into three discrete risk levels: high risk (red), medium risk (yellow), and low risk (green). The discrete risk level labels are time-aligned and fused with the incremental cost trend data. According to the engineering cost report template, a project cost analysis report with risk level color block annotations, cost deviation curves, and risk evolution stage descriptions is generated.

[0104] It should also be explained that the incremental cost trend is obtained by calculating the column mean of a multidimensional cost feature matrix. A three-dimensional matrix is ​​formed after parsing multiple standardized visa document packages. The arithmetic mean of the values ​​in each column of the second dimension of the matrix is ​​calculated and output as an incremental cost trend of length N.

[0105] This embodiment also provides a project cost accounting management system for engineering supervision, comprising: a data acquisition module for collecting raw data from engineering monitoring, performing signal strength screening and timestamp synchronization processing, and generating pre-processed data packets;

[0106] The edge cleaning module is used to input pre-processed data packets into the edge computing node, perform dynamic data cleaning using a noise filtering matrix algorithm, and output a structured data set with quality labels;

[0107] The verification and evidence storage module is used to scan the construction site using augmented reality glasses based on a structured data set with quality labels, and compare the actual coordinates with the design coordinates of the building information model in real time. When the accumulated deviation value exceeds the warning threshold, an unalterable acceptance evidence chain is generated;

[0108] The visa generation module is used to automatically generate a standardized visa document package based on the deviation data in the acceptance evidence chain and calculate the incremental cost using the compensation coefficient matrix;

[0109] The risk prediction module is used to accumulate the incremental costs of multiple standardized visa document packages, perform time series analysis through long-short-term memory prediction models, and output a project cost analysis report with risk levels.

[0110] This embodiment also provides a computer device suitable for the project cost accounting management method for engineering supervision, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the project cost accounting management method for engineering supervision proposed in the above embodiment.

[0111] The computer device may be a terminal, comprising a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and computer programs. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The communication interface of the computer device is used to communicate with external terminals via wired or wireless communication. Wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display. The input device may be a touchscreen overlay on the display, buttons, a trackball, or a touchpad on the computer device housing, or an external keyboard, touchpad, or mouse.

[0112] This embodiment also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the project cost accounting management method for engineering supervision proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0113] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A project cost accounting management method for engineering supervision, characterized in that: The following steps are involved: S1. Collect raw data from engineering monitoring, perform signal strength screening and timestamp synchronization processing, and generate pre-processed data packets; S2: Input the pre-processed data packet into the edge computing node, use the noise filtering matrix algorithm to perform dynamic data cleaning, and output a structured data set with quality labels; S3. Based on a structured data set with quality labels, the construction site is scanned using augmented reality glasses, and the actual coordinates are compared with the design coordinates of the building information model in real time. When the accumulated deviation exceeds the warning threshold, an unalterable acceptance evidence chain is generated; S4. Based on the deviation data in the acceptance evidence chain, the incremental cost is calculated using the compensation coefficient matrix and a standardized visa document package is automatically generated; S5. Accumulate the incremental costs of multiple standardized visa document packages, perform time series analysis using a long-short-term memory prediction model, and output a project cost analysis report with risk levels. Among them, in S2, the output is a structured data set with quality labels, including: Perform quantum chromodynamics noise modeling on the preprocessed data packets, output quantum noise characteristic tensors, and construct turbulent fluid dynamics equations to solve the filtered data field; By integrating the chaotic pulse synchronization mechanism with Lie group differential manifolds, the multi-source time domain alignment of the filtered data field is performed, and the fractional-order Kalman filter algorithm is used to eliminate the drift error of the filtered data field after time domain alignment to generate drift-suppressed data. A dynamic weight network is constructed based on the spatial topological relationship of drift suppression data. The consistency of the associated data sets between the spatial topological associated nodes is checked to generate discrete quality grade labels. These labels are then fused with the drift suppression data to output a structured data set with quality labels. In S3, an unalterable acceptance evidence chain is generated, including: A quantum random number generator is used to quantize and encode the three-dimensional coordinate vectors in a structured data set with quality labels to generate quantum entangled state data. The laser radar of augmented reality glasses is used to scan the construction site to capture point cloud data in real time. Real-time temperature data from the construction site is collected and the cumulative deviation of the thermodynamic entropy constraint is calculated. When the cumulative deviation exceeds the warning threshold, the point cloud topology genus features of the point cloud data are extracted, a three-dimensional feature hash value is generated, and a spatial topology certificate is signed using a quantum key. The three-dimensional feature hash value and spatial topology certificate are submitted to the alliance chain, and a verifiable delay function is performed on the public chain through the cross-chain oracle to generate zero-knowledge proof and build an unalterable acceptance evidence chain.

2. A project cost accounting management method for engineering supervision according to claim 1, characterized in that: In S1, the collected engineering monitoring raw data include spatial positioning data, structural deformation data, material performance parameters and environmental status parameters.

3. A project cost accounting management method for engineering supervision according to claim 2, characterized in that: The signal strength screening and timestamp synchronization processing are performed to generate a pre-processed data packet, including: The signal strength and environmental noise level of each sensor node are extracted from the engineering monitoring raw data, and the screening score is calculated using a dynamic adaptive threshold model. The corresponding engineering monitoring raw data with a score below the dynamic threshold screening score is marked as invalid and discarded. Use the reference clock node to compare with other node clocks, calculate the time offset, apply the weighted correction model to adjust the timestamps of all engineering monitoring raw data, apply the final integration formula to unify the format of engineering monitoring raw data, and output the preprocessed data packet.

4. A project cost accounting management method for engineering supervision according to claim 1, characterized in that: In S4, a standardized visa document package is automatically generated, including: The deviation accumulation value and cloud topology defect characteristics are extracted from the acceptance evidence chain and encoded. After quantum signature verification, the deviation data is output and input into the quantum coprocessor for variational quantum feature solution to obtain the compensation eigenvalue. Performing a Hadamard product operation on the compensation eigenvalue and the preset compensation coefficient tensor to obtain the incremental cost; Call the distributed legal engine to load the standard engineering contract template library, take the incremental cost and deviation characteristics as input parameters, apply the Laplace operator constraints of the contract terms to perform least squares optimization solution, and output a standardized visa document package.

5. A project cost accounting management method for engineering supervision according to claim 1, characterized in that: In S5, a project cost analysis report with risk levels is output, including: Decode the incremental cost characteristics of multiple standardized visa document packages and output a multi-dimensional cost characteristic matrix; The multi-dimensional cost feature matrix is ​​input into the molecular dynamics simulation engine for continuous homology analysis to generate a dynamic topological feature sequence; The dynamic topological feature sequence is input into the long-short-term memory prediction model to predict the evolution in the time dimension, output the incremental cost risk time series prediction data, and inject quantum random perturbations to output the risk evolution path; Perform spatial convergence analysis on the risk evolution path to obtain the dynamic state characteristic vector, and perform discrete risk level division. This is then integrated with the incremental cost trend to generate a project cost analysis report with risk levels.

6. A project cost accounting management system for engineering supervision, based on the project cost accounting management method for engineering supervision according to any one of claims 1 to 5, characterized in that: include, The data acquisition module is used to collect raw data from engineering monitoring, perform signal strength screening and timestamp synchronization processing, and generate pre-processed data packets; The edge cleaning module is used to input pre-processed data packets into the edge computing node, perform dynamic data cleaning using a noise filtering matrix algorithm, and output a structured data set with quality labels; The verification and evidence storage module is used to scan the construction site using augmented reality glasses based on a structured data set with quality labels, and compare the actual coordinates with the design coordinates of the building information model in real time. When the accumulated deviation value exceeds the warning threshold, an unalterable acceptance evidence chain is generated; The visa generation module is used to automatically generate a standardized visa document package based on the deviation data in the acceptance evidence chain and calculate the incremental cost using the compensation coefficient matrix; The risk prediction module is used to accumulate the incremental costs of multiple standardized visa document packages, perform time series analysis through long-short-term memory prediction models, and output a project cost analysis report with risk levels.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the project cost accounting management method for engineering supervision according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the project cost accounting management method for engineering supervision according to any one of claims 1 to 5 are implemented.

Images