Coal construction project supervision and management method and system based on big data

Abstract

The invention relates to the technical field of coal engineering data monitoring, in particular to a coal construction engineering supervision and management method and system based on big data, and the method comprises the steps: enabling consumed materials and engineering progress to form data points in the construction process of the coal construction engineering, and setting a first neighborhood range of the data points, and determining the neighborhood anomaly degree according to the fluctuation condition of the first neighborhood range of the data point, and adjusting the first neighborhood range to obtain a second neighborhood range. And evaluating the adaptation degree of the second neighborhood range, and calculating a third neighborhood range. And determining robust weights of the data points according to the third neighborhood range, and performing least square fitting based on the robust weights of all the data points to obtain an overall fitting curve. The slope change of the overall fitting curve is analyzed, and whether the data points of the current date are abnormal or not is judged, so that real-time supervision and management are achieved. According to the method, the abnormal relationship between the consumed materials and the project progress in the construction process can be effectively identified, and the timeliness and accuracy of construction supervision and management are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of coal engineering data monitoring, and in particular to a coal construction engineering supervision and management method and system based on big data. Background Art

[0002] In the field of coal construction projects, the projects are large-scale and have long construction periods. Abnormal monitoring of project progress and material consumption can achieve refined control of the project construction process.

[0003] Robust weighted least squares is a common technical means. Its specific implementation method is: based on historical construction data and multi-source information obtained by real-time monitoring, a correlation model between material consumption and project progress is constructed. First, for each data point, other related data points are collected within its preset fixed neighborhood. Then, the robust weight of the current data point is determined based on the degree of influence of each data point in the neighborhood on the robust weight of the current data point. The least squares method is used to fit these weighted data points to generate a fitting curve that reflects the relationship between material consumption and project progress. During the progress of the project, abnormal monitoring is performed based on the changes in the fitting curve.

[0004] However, in the complex and changeable actual scenario of coal construction projects, the traditional robust weighted least squares method has obvious defects. The construction environment of coal construction projects is complex, and the construction process is affected by many factors such as geological conditions, climatic factors, and technical process adjustments. In this context, it is difficult for a fixed neighborhood range to flexibly respond to changes in local characteristics of different data points. If the neighborhood range is too small, the model can only capture limited local data changes, and excessive focus on local features will lead to misjudgment of the overall trend of material consumption and project progress, and cannot accurately reflect the actual situation of the project. If the neighborhood range is set too large, the calculation efficiency will be seriously reduced, delaying the best time to solve the problem.

[0005] In summary, when using the traditional robust weighted least squares method to monitor the progress and material consumption of coal construction projects, the fixed neighborhood range will affect the accuracy of the abnormal monitoring results, leading to misjudgment of the progress and material consumption of the project, and cannot meet the accurate supervision and management of coal construction projects. Summary of the invention

[0006] In order to solve the problem that when using the traditional robust weighted least squares method to perform abnormal monitoring of project progress and material consumption, the fixed neighborhood range will affect the accuracy of the abnormal monitoring results and cannot meet the accurate supervision management, the present invention proposes a coal construction project supervision management method and system based on big data.

[0007] In a first aspect, the present invention provides a coal construction project supervision and management method based on big data, comprising: During the construction of the coal construction project, the total amount of materials consumed and the progress of the project each day constitute a data point, all data points from the start of construction to the current date are obtained, and a two-dimensional coordinate system of all data points is constructed with the total amount of materials consumed as the horizontal axis and the progress of the project as the vertical axis; Presetting the first neighborhood range of all data points; determining the neighborhood anomaly degree of each data point according to the fluctuation degree of the first neighborhood range of each data point, and adjusting the first neighborhood range of each data point to the second neighborhood range using the neighborhood anomaly degree of each data point; Evaluate the fitness of the second neighborhood range of each data point and calculate the third neighborhood range of each data point: , in the formula, For the The third neighborhood of data points, For the The degree of adaptation of the second neighborhood range of the data point, is the preset first neighborhood range, For the The second neighborhood range of data points; The robust weight of each data point is determined according to the third neighborhood range of each data point, and the least squares fitting is performed based on the robust weights of all data points to obtain the overall fitting curve. The material consumption and project progress of the current date are supervised and managed in real time according to the change in the slope of the overall fitting curve.

[0008] This technical solution first combines the total amount of materials consumed and the project progress every day into one data point, constructs a basic data unit that reflects the relationship between project progress and material consumption, and obtains comprehensive data points with time series characteristics. These data can intuitively reflect the changes in the correlation between material consumption and project progress as the project progresses, and can more accurately grasp the actual progress of the project. Then, the first neighborhood range is dynamically adjusted according to the degree of neighborhood anomaly of each data point, so that the second neighborhood range can adapt to the local characteristics of different data points, avoiding the limitations of the fixed first neighborhood range. Then, an indicator of the degree of adaptation to measure the quality of the second neighborhood range is introduced, and the second neighborhood range is further optimized by comprehensively considering multiple factors to obtain the third neighborhood range of each data point. The third neighborhood range can better balance the local characteristics and overall trends of the data points, making the subsequent analysis based on the third neighborhood range more accurate and reliable, and improving the modeling accuracy of the relationship between project progress and material consumption. Finally, the robust weights are determined according to the third neighborhood range, so that the calculation of the robust weights can more accurately reflect the importance of the data points. The accurate robust weight calculation and the overall fitting curve obtained by least squares fitting can more accurately describe the relationship between material consumption and project progress. By monitoring the slope change of the overall fitting curve to judge the anomaly, the abnormal changes in project progress and material consumption can be discovered in time, and real-time supervision and management of coal construction projects can be achieved, so that timely measures can be taken to make adjustments, ensure the smooth progress of the project, and improve the efficiency and quality of project management.

[0009] Furthermore, curve fitting is performed on all data points within the third neighborhood of each data point to obtain a local curve for each data point, and the corresponding The coordinate value and the original The difference in coordinate values, and the reciprocal of the absolute value of the difference is taken as the robust weight of the data point.

[0010] The robust weight of each data point determined by this technical solution makes the robust weights of data points with similar changing trends to the surrounding data points relatively large. When fitting the overall fitting curve of the relationship between material consumption and project progress, these data points with large robust weights play a leading role in the fitting process of the overall curve, so that the final overall fitting curve can more accurately reflect the true relationship between material consumption and project progress, and provide a reliable basis for the subsequent accurate judgment of whether the project progress and material consumption are abnormal.

[0011] Furthermore, the degree of adaptation of the second neighborhood range of each data point satisfies the following relationship: , , For the The degree of adaptation of the second neighborhood of the data point and the total number of data points in the second neighborhood, , For the data points, The second neighborhood of the data point The local fitting error of the data points is , For the data points, The second neighborhood of the data point The degree of abnormality of the neighborhood of a data point, , For the The standard deviation of the local fitting error and the standard deviation of the neighborhood anomaly of all data points within the second neighborhood of the data point, , For the The mean of the local fitting errors and the mean of the neighborhood anomaly of all data points within the second neighborhood of a data point, is the natural exponential function, is the absolute value symbol.

[0012] This technical solution comprehensively considers multi-dimensional data features to determine the degree of adaptation of the second neighborhood range, comprehensively weighs the local fitting error, neighborhood anomaly degree and discrete characteristics of each data point within the second neighborhood range, and can accurately measure the quality of the second neighborhood range, providing reliable analysis indicators for the subsequent determination of a more accurate third neighborhood range.

[0013] Furthermore, the method for real-time supervision and management of the material consumption and engineering progress on the current date according to the change in the slope of the overall fitting curve is: The slope at each data point on the overall fitting curve is obtained, and all slopes are detected for anomalies using the box plot method; if the slope at the data point of the current date is abnormal, it is determined that the material consumption and project progress of the current date are abnormal, and an early warning notification is issued, and the consumed materials and project progress of the current date are checked and analyzed; if the slope at the data point of the current date is not abnormal, it is determined that the material consumption and project progress of the current date are not abnormal, and no early warning notification is issued.

[0014] Furthermore, the first neighborhood range, the second neighborhood range, and the third neighborhood range of each data point are all length values, and each data point is used as a starting point, and the neighborhood range corresponding to the data point is obtained by intercepting forward according to the length value.

[0015] Furthermore, the neighborhood anomaly degree of each data point is calculated based on the following formula: ; In the formula, For the The degree of abnormality of the neighborhood of a data point, For the The fluctuation degree of the first neighborhood range of the data point, For the The total number of data points within the first neighborhood of a data point, For the The first neighborhood of the data point The original data point Coordinate values, For the The first neighborhood of the data point The original data point Coordinate values, For the The original data points of all data points in the first neighborhood of the data point The mean of the coordinate values, is the preset first neighborhood range, is the absolute value symbol, is a natural exponential function.

[0016] This technical solution calculates the neighborhood anomaly degree of each data point and can accurately reflect the abnormal situation of the data point within its neighborhood range. The accurate anomaly degree assessment provides a reliable basis for the subsequent adjustment of the first neighborhood range, so that the adjusted second neighborhood range can better adapt to the local characteristics of the data point, thereby improving the accuracy of data analysis.

[0017] Furthermore, the local fitting error is determined based on the following method: curve fitting is performed on all data points within the second neighborhood of each data point to obtain a local curve for each data point; the corresponding error of the data point on the local curve is calculated. The coordinate value and the original The absolute value of the difference in coordinate values ​​is taken as the local fitting error of the data point.

[0018] Furthermore, the fluctuation degree of the first neighborhood range of each data point is determined based on the following method: calculating the fluctuation degree of all data points within the first neighborhood range of each data point Variance of coordinate values , and all data points in the coordinate system Variance of coordinate values ;Will and The ratio of is taken as the fluctuation degree of the first neighborhood range of the data point.

[0019] This technical solution quantifies the degree of discreteness of all data points within the first neighborhood of each data point by calculating the variance of all data points within the first neighborhood. It can intuitively reflect the fluctuation of all data points within the first neighborhood. This quantification method can more accurately understand the stability of data points within the neighborhood of each data point.

[0020] Furthermore, the method of adjusting the first neighborhood range of each data point to the second neighborhood range by using the neighborhood abnormality degree of each data point is as follows: preset the neighborhood abnormality degree threshold ; Calculate the second neighborhood range for each data point: , where For the The second neighborhood of data points, is the preset first neighborhood range, For the The degree of abnormality of the neighborhood of a data point, The round-up symbol.

[0021] This technical solution establishes a clear correlation between the degree of neighborhood anomaly and the first neighborhood range. The degree of neighborhood anomaly of each data point accurately reflects the abnormal situation around the data point. Compared with the traditional method of fixing the neighborhood range, this solution can flexibly and dynamically adjust the first neighborhood range to a more appropriate second neighborhood range according to the actual abnormal situation of the first neighborhood range of each data point.

[0022] In a second aspect, the present invention provides a coal construction project supervision and management system based on big data, the supervision and management system comprising a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement any step of the supervision and management method.

[0023] The present invention has the following effects: This scheme obtains data points with time series characteristics covering the entire construction process, first preliminarily adjusts the neighborhood range according to the data distribution characteristics of the neighborhood range of the data points, and then introduces the adaptation degree index to adjust and optimize the neighborhood range again, balancing the local details and overall trends of the data points. The least squares method is used based on the robust weights determined based on the optimized neighborhood range to generate an overall fitting curve that accurately reflects material consumption and project progress, thereby improving the fitting effect. Based on the precise overall fitting curve, accurate abnormality monitoring can be carried out, realizing real-time and efficient supervision and management of coal construction projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic flow chart of the method of the present invention; Figure 2 is the overall fitting curve diagram of the present invention; Figure 3 It is a flow chart of the present invention for real-time supervision and management of material consumption and engineering progress on the current date based on the overall fitting curve; Figure 4 It is a box plot anomaly detection result diagram of the present invention. DETAILED DESCRIPTION

[0025] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0026] Reference Figure 1 The present invention provides a coal construction project supervision and management method based on big data, comprising steps S1 to S7: S1: Collect material consumption and project progress, and construct a two-dimensional coordinate system.

[0027] During the construction of coal construction projects, the daily material consumption is huge, among which concrete, as a key building material, plays an important role in the progress of the project. In the supervision and management system involved in the present invention, the consumed materials mentioned refer specifically to the total amount of concrete consumed.

[0028] In order to accurately grasp the dynamics of the project, data collection work is carried out on a regular basis every day. Specifically, the total amount of consumed materials and project progress data are collected once a day. For example, when data is collected on the 10th day, the total amount of consumed materials obtained is the cumulative consumption of concrete from the start date of the project (day 1) to the 10th day; and the project progress data collected on the 10th day reflects the actual progress of the project as of the 10th day. Similarly, the total amount of consumed materials collected on the 11th day is the total consumption of concrete from the 1st day to the 11th day when the project started, and the project progress collected on the 11th day represents the progress status of the project on the 11th day.

[0029] The total amount of materials consumed and the progress of the project are collected every day from the beginning of construction to the current date. The total amount of materials consumed is used as the horizontal axis and the progress of the project is used as the vertical axis to construct a two-dimensional coordinate system covering all data points. In this way, the total amount of materials consumed and the progress of the project collected daily together constitute a data point in the two-dimensional coordinate system.

[0030] In the two-dimensional coordinate system, The horizontal coordinate of the data point , indicating the The total amount of materials consumed on the date corresponding to the data point. The vertical coordinate of the data point , indicating the The data point corresponds to the project progress on the date.

[0031] In short, by constructing a two-dimensional coordinate system with the total amount of consumed materials as the horizontal axis and the project progress as the vertical axis, the complex project progress and material consumption can be presented intuitively. Each data point represents the construction status of a day. The distribution and change trend of the overall data points allow managers to understand the progress trajectory of the project from the beginning to the present at a glance, providing a solid foundation for real-time supervision and management.

[0032] S2: Determine the degree of neighborhood anomaly based on the first neighborhood range of each data point.

[0033] In the supervision and management system of the present invention, the degree of neighborhood abnormality of each data point is determined based on the first neighborhood range of each data point. According to the traditional robust weighted least squares method, a fixed neighborhood range is set for each data point. is 7 (empirical value), and the size of the first neighborhood range of all data points is .in, is a length value, for example , for the For each data point, data point as the starting point, intercept 5 data points forward, and data points and 5 intercepted data points as the The first neighborhood of data points.

[0034] In the process of supervision and management of coal construction projects, it is crucial to accurately grasp the characteristics and laws of data to ensure the smooth progress of the project. If the stability of the data points within the first neighborhood of a data point is poor and the outlier is high, it means that the neighborhood abnormality of the data point is greater. If it is not adjusted and the robust weights are still obtained according to the first neighborhood range, the results will be difficult to accurately reflect the actual situation, which will affect the accurate judgment of the project status.

[0035] First, analyze the fluctuation degree of the first neighborhood range of each data point: Calculate the first neighborhood of each data point. Variance of coordinate values , and all data points in the coordinate system Variance of coordinate values ;Will and The ratio of is used as the fluctuation degree of the first neighborhood range of the data point to effectively evaluate the stability of the data in the local area where the data point is located. The larger the ratio, the greater the fluctuation degree of the first neighborhood range of the data point, and the more unstable the data point in the first neighborhood range, and vice versa.

[0036] For data points, the fluctuation degree of the first neighborhood range is:

[0037] In this formula, For the The fluctuation degree of the first neighborhood range of the data point, For the All data points within the first neighborhood of a data point The variance of the coordinate values, is the coordinate system for all data points Variance of the coordinate values.

[0038] Variance is a statistic that measures the degree of data dispersion and can directly reflect the fluctuation of data. The greater the fluctuation of the first neighborhood of a data point, the greater the Relative to The larger the The fluctuation of the data point in the first neighborhood of the data point is more intense than the overall fluctuation, indicating that the data point in the first neighborhood of the data point is unstable, which means that in the construction stage corresponding to the first neighborhood, there are factors that affect the stability of the project progress, such as construction process adjustment, equipment failure or personnel deployment problems. The smaller the fluctuation degree of the first neighborhood range of a data point, the more it means that the fluctuation degree of the data point in the first neighborhood range of the data point is within the normal range of the overall fluctuation, and the data points in the first neighborhood range of the data point are relatively stable, which means that in the construction stage corresponding to the first neighborhood range, the project is in a stable progress state.

[0039] Then, calculate the degree of neighborhood anomaly for each data point:

[0040] In the formula, For the The degree of neighborhood anomaly of a data point comprehensively reflects the anomaly of all data points within the first neighborhood of the data point. The larger the value, the higher the degree of anomaly of the first neighborhood of the data point and the more unstable the local data. For the The degree of fluctuation of the first neighborhood range of a data point reflects the magnitude of the fluctuation of the first neighborhood range of the data point relative to the fluctuation of the overall data. The larger the value, the more drastic the fluctuation of the first neighborhood of the data point is compared with the overall value. It will amplify the impact of other factors on the degree of anomaly, emphasize the contribution of neighborhood fluctuations to the overall degree of anomaly, and make the final result more prominent in the unstable characteristics of the first neighborhood range. For the The total number of data points within the first neighborhood of a data point, For the The first neighborhood of the data point The original data point Coordinate values, For the The first neighborhood of the data point The original data point Coordinate values, For the The original data points of all data points in the first neighborhood of the data point The mean of the coordinate values, is the preset first neighborhood range, is the absolute value symbol, is a natural exponential function. The coordinate value indicates the total amount of material consumed. The coordinate value is the project progress.

[0041] In this formula, Combined with the The original data points of all data points in the first neighborhood of the data point The coordinate value is relative to the original The deviation from the average level of the coordinate values, and The distance weight between a data point and all the data points in its first neighborhood in the dimension of the total amount of consumed materials can accurately reflect the deviation of all the data points in the first neighborhood. The department reflects The deviation of the engineering progress of each data point within the first neighborhood of the data points from the average engineering progress. The larger the deviation, the higher the degree of outlier. Part is the distance weight calculated by the Gaussian kernel function, Measured the data point and the first neighborhood The distance between data points on the total amount of consumed material, according to The principle is to use the first neighbor range Instead of the standard deviation in the Gaussian kernel function, it is used to adjust the width of the Gaussian kernel function. When the distance between two data points in the total material consumption is closer, the value of this part is closer to 1, which means that The data points in the first neighborhood of the data point are The greater the influence of the abnormal degree of the neighborhood of the data point, the greater the impact; on the contrary, the farther the distance, the closer the value of this part is to 0, which means that the The data points in the first neighborhood of the data point are This distance weighting method takes into account the correlation of data points in the dimension of the total amount of consumed materials, and more reasonably integrates the contribution of data points at different positions within the first neighborhood to the neighborhood anomaly of the current data point, so that the calculation results can more accurately reflect the first The actual situation of the data points within the first neighborhood of the data points.

[0042] In summary, this formula comprehensively and meticulously reflects the abnormal conditions of the neighborhood of each data point by associating the neighborhood fluctuations with the overall fluctuations of the data point, combining the degree to which the data points in the neighborhood deviate from the mean in terms of engineering progress, and the distance weights of the data points in the neighborhood in terms of the total amount of consumed materials.

[0043] S3: Adjust the first neighborhood range to the second neighborhood range according to the degree of neighborhood anomaly.

[0044] If the degree of neighborhood anomaly of a data point is greater, it means that the uncertainty of the data points within the first neighborhood of the data point is higher. At this time, a larger neighborhood range is needed to integrate more data information so that the robust weights obtained by the calculation tend to be stable, and vice versa.

[0045] Therefore, the first neighborhood range is adjusted to the second neighborhood range as follows: Preset neighborhood anomaly threshold It is 0.47 (experience value).

[0046] Compute the second neighborhood of each data point:

[0047] In this formula, For the The second neighborhood of data points, is the preset first neighborhood range, For the The degree of abnormality of the neighborhood of a data point, The round-up symbol.

[0048] like Greater than , indicating that the abnormality of the neighborhood of the data point exceeds the preset normal range, and the neighborhood range needs to be expanded to stabilize the robust weight. Greater than 1, for the preset first neighborhood range To enlarge; if Less than , it means that the abnormality of the neighborhood is relatively low, and the adjustment range of the neighborhood range is correspondingly small. Less than 1, for the preset first neighborhood range Zoom out.

[0049] This adjustment strategy can adapt to the complex and changeable data characteristics in coal construction projects. By expanding or reducing the neighborhood range and integrating more or more accurate data information, the calculated robust weights tend to be stable, providing a more reliable data basis for subsequent least squares fitting based on robust weights and engineering supervision management decisions.

[0050] in, is a length value, for example , then take data point as the starting point, intercept 10 data points forward, and data points and 10 intercepted data points as the The second neighborhood of data points.

[0051] S4: Evaluate the degree of adaptation of the second neighborhood range.

[0052] Accurately assessing the status of a project depends on accurately understanding the relationship between material consumption and project progress data. The degree of adaptation of the second neighborhood range helps to identify neighborhood data that can more accurately reflect the actual progress of the project. For example, when the degree of adaptation of the second neighborhood range of a data point is high, it means that the data points in the neighborhood are more consistent and reliable in reflecting the relationship between material consumption and project progress. Based on this, managers can more accurately judge whether the actual progress of the project at this stage is in line with expectations, whether the material consumption is reasonable, and promptly discover potential project problems, such as delayed progress or material waste.

[0053] When the robust weight of each data point is subsequently calculated, the second neighborhood range with a high degree of adaptation can provide a more reliable data basis, so that the robust weight calculated based on these data can more accurately reflect the true distribution of the data, avoiding the bias of the robust weight due to improper selection of the neighborhood range, including too much abnormal or unstable data, and thus affecting the accuracy of subsequent data processing results such as least squares fitting.

[0054] First, get the local fitting error for each data point: Perform curve fitting on all data points within the second neighborhood of each data point to obtain the local curve of each data point; calculate the corresponding local curve of the data point The coordinate value and the original The absolute value of the difference in coordinate values ​​is taken as the local fitting error of the data point.

[0055] Next, the degree of adaptation of the second neighborhood range of each data point is calculated according to the following formula:

[0056] In this formula, For the The degree of adaptation of the second neighborhood range of a data point directly reflects the desirability of the neighborhood range in overall data processing and engineering analysis. The larger the value, the more reliable and stable the data points covered by the second neighborhood range are in reflecting key information such as the relationship between material consumption and engineering progress, and the more suitable they are for subsequent robust weight calculations. For the The total number of data points within the second neighborhood of data points, For the The local fitting error of the data points is For the The second neighborhood of the data point The local fitting error of the data points is For the The degree of abnormality of the neighborhood of a data point, For the The second neighborhood of the data point The degree of abnormality of the neighborhood of a data point, For the The standard deviation of the local fitting error for all data points within the second neighborhood of data points, For the The standard deviation of the neighborhood anomaly of all data points within the second neighborhood of a data point, For the The mean of the neighborhood abnormality of all data points within the second neighborhood of a data point, For the The mean of the local fitting errors of all data points in the second neighborhood of the data point, is the natural exponential function, is the absolute value symbol.

[0057] In this formula, multiply by It plays a normalization role. Since the number of data points contained in the second neighborhood range of different data points may vary greatly, if normalization is not performed, a data point with many data points in its second neighborhood range may dominate the cumulative calculation, while a data point with few data points in its second neighborhood range may have little impact on the cumulative calculation, making it impossible to fairly compare the degree of adaptation of the second neighborhood range of different data points. Through normalization, the calculation results of different data points are more comparable at the same scale.

[0058] In this formula, part, Measured the The local fitting error of a data point is related to its second neighborhood. The relative difference between the local fitting errors of the data points reflects the The relative change in fitting error between a data point and each data point in its second neighborhood; Measured the The abnormality of the neighborhood of a data point is related to the abnormality of the second neighborhood. The relative difference in the degree of abnormality of the data points, and It is used for normalization and 1 is added to avoid the generation of zero or negative values, so as to more stably reflect the degree of adaptation of the second neighborhood range.

[0059] This part comprehensively considers the coordinated changes of each data point and the data points in its second neighborhood in two key dimensions: local fitting error and neighborhood anomaly degree. The larger the calculation result, the more consistent the changes of each pixel point and the data points in its second neighborhood in these two dimensions. The correlation between the data points in the second neighborhood is strong, which is more in line with the inherent laws of engineering data, and the adaptation degree of the second neighborhood is higher. On the contrary, the smaller the calculation result, the more chaotic the changes of the data points in the second neighborhood in these two dimensions, the lower the quality of the second neighborhood, and the lower the adaptation degree.

[0060] In this formula, part, It reflects the The ratio of the local fitting error of a data point to the abnormality of the neighborhood reflects the The relative relationship of each data point itself on these two key factors. Indicates The ratio of the mean of the local fitting errors of all data points within the second neighborhood of a data point to the mean of the neighborhood anomaly of all data points reflects the relative relationship between these two factors for all data points within the second neighborhood. Weighed and The smaller the difference, the better the consistency of the two relative relationships, indicating that The greater the correlation between the local fitting error and the degree of neighborhood anomaly of each data point in its second neighborhood, the greater the correlation between the local fitting error and the degree of neighborhood anomaly of each data point in its second neighborhood. The more data points in the second neighborhood of a data point conform to the rule, the better the The greater the degree of adaptation of the second neighborhood range of the data point, the greater the degree of adaptation of the second neighborhood range of the data point. Build and negative correlation, and vice versa.

[0061] In summary, by comprehensively considering the relationship between the local fitting error of the data points in the second neighborhood of each data point and the degree of neighborhood anomaly, as well as the consistency of each data point with the data points in the second neighborhood in key features, the degree of adaptation of the second neighborhood of each data point can be accurately evaluated. The higher the degree of adaptation, the more effectively the second neighborhood can screen out high-quality neighborhood data.

[0062] S5: Comprehensively determine the third neighborhood range of each data point based on the first neighborhood range, the second neighborhood range and the degree of adaptation of each data point.

[0063] As of the current step, the first neighborhood range and the second neighborhood range of each data point have been obtained. The first neighborhood range is a preset initial range, which provides a basic range for data processing, but cannot fully adapt to the specific situation of each data point. The second neighborhood range is an appropriate adjustment of the first neighborhood range, but its reliability and applicability also vary, and need to be measured by the degree of adaptation.

[0064] Since the characteristics and environments of different data points are different, it is difficult to accurately capture the real characteristics and internal laws of the data by using only the first neighborhood range or the second neighborhood range. In order to more accurately reflect the relationship between the data point and the surrounding data and improve the accuracy and reliability of data processing, this step comprehensively considers the first neighborhood range, the second neighborhood range and their adaptability, and determines a reasonable third neighborhood range, so that subsequent operations such as robust weight calculation based on this range are more accurate and reliable.

[0065] Specifically, the second neighborhood range is adjusted to the third neighborhood range according to the following relationship:

[0066] In this formula, For the The third neighborhood of data points, For the The degree of adaptation of the second neighborhood range of the data point, is the preset first neighborhood range, For the The second neighborhood of data points.

[0067] In this formula, when The larger the value, the The second neighborhood range of the data point is more reliable and stable in reflecting key information such as the relationship between material consumption and project progress, and is more suitable for subsequent calculations. This means that the second neighborhood range will be used more frequently. To calculate the third neighborhood range (the final neighborhood range), because this neighborhood range is more desirable. On the contrary, when When it is smaller, it means that the reliability of the second neighborhood is low. It will be relatively large, so the calculation will rely more on the preset first neighborhood range. , that is, reducing the reliance on the unreliable second neighborhood range.

[0068] in, is a length value, for example , then take data point as the starting point, intercept 20 data points forward, and data points and 20 intercepted data points as the The third neighborhood range of data points.

[0069] S6: Determine the robust weight of each data point through the third neighborhood range.

[0070] Curve fitting is performed on all data points within the third neighborhood of each data point to obtain the local curve of each data point. Because data points of material consumption and engineering progress data that are close in space or time often have similar change trends, this correlation can be fully explored by using the data in the neighborhood for curve fitting, making the fitted local curve more consistent with the actual data distribution, thereby more accurately reflecting the inherent laws of the data.

[0071] Calculate the difference between the Y coordinate value corresponding to the data point on the local curve and the original Y coordinate value of the data point, and take the reciprocal of the absolute value of the difference as the robust weight of the data point. This operation can intuitively reflect the degree of fit between the data point and the local curve. The smaller the difference, the closer the data point is to the local curve, which means that the consistency of its change trend with other data points in the neighborhood is higher, and the larger the robust weight obtained by taking the reciprocal. On the contrary, the larger the difference, the smaller the robust weight. For example, in a coal construction project, if the robust weight of a data point is large, it means that the relationship between the material consumption and engineering progress represented by the data point is consistent with the overall situation in the neighborhood and is more trustworthy; while data points with smaller robust weights may have abnormalities and require further attention and analysis.

[0072] This robust weight calculation method can highlight those data points that fit the local curve well. In subsequent data processing and analysis, such as least squares fitting, data points with large robust weights will have a greater impact on the results, which can make the analysis results more inclined to reliable data and improve the accuracy and stability of overall data processing. Taking engineering progress prediction as an example, the use of this robust weight calculation method can make the prediction results more dependent on reliable data points, reduce the interference of abnormal data on the prediction results, and thus improve the accuracy of the prediction.

[0073] S7: Least squares fitting is performed based on robust weights, and real-time supervision and management of material consumption and project progress on the current date is performed based on the overall fitting curve.

[0074] In real-time supervision and management, it is crucial to accurately identify anomalies in material consumption and project progress. By setting reasonable robust weights for data points, the fitting curve can more accurately reflect the characteristics of engineering data under normal circumstances. When the slope of a data point on a certain date is abnormal, since the fitting curve has fully considered the importance of data at different stages, it can be more confidently determined that this is a real abnormality rather than a misjudgment caused by unreasonable data weights.

[0075] Therefore, the robust weight of each data point is used as the weight for least squares fitting, and then the fitting operation is performed on all data points (all data points collected from the beginning of the project to the current date), and finally an overall fitting curve is obtained, such as Figure 2 as shown in .

[0076] Subsequently, the process of real-time supervision and management of the material consumption and project progress on the current date based on the overall fitting curve is as follows: Figure 3 As shown: S71: Obtain all slopes on the overall fitting curve.

[0077] After obtaining the overall fitting curve, the slope at each data point on the curve is calculated to obtain all slope data.

[0078] S72: Use the box plot method to perform anomaly detection on all the slope data obtained. The results are as follows: Figure 4 As shown in the figure, the two dotted lines in the middle are the normal range of the slope, and the slope on the far left exceeds the normal range. This slope is an abnormal slope. Compare the slope of the data point on the current date with the normal range of the slope to determine whether it is abnormal.

[0079] S73: Determine whether the slope at the data point corresponding to the current date is abnormal.

[0080] If an abnormal situation occurs, execute step S731: determine that the material consumption and project progress on the current date are abnormal, and then execute step S7311: issue an early warning notification, and conduct a detailed inspection and analysis of the material situation consumed on the current date and the project progress situation.

[0081] If no abnormal situation occurs, execute step S732: determine that the material consumption and project progress on the current date are in a normal state, and then execute step S7321: do not issue a warning notification.

[0082] Through the above operation process, real-time supervision and management of material consumption and project progress on the current date is realized. For each day in the construction process, the overall fitting curve of all data points can be obtained according to the process described in the present invention, and according to the change of the slope of the overall fitting curve, real-time and accurate supervision and management of material consumption and project progress on each day can be implemented.

[0083] The present invention also provides a coal construction project supervision and management system based on big data, the supervision and management system includes a memory and a processor, the memory stores a computer program, the processor executes the computer program to implement the steps of any supervision and management method, and realizes the supervision and management of material consumption and project progress.

Claims

1. A coal construction project supervision and management method based on big data, characterized in that: include: During the construction of the coal construction project, the total amount of materials consumed and the progress of the project each day constitute a data point, all data points from the start of construction to the current date are obtained, and a two-dimensional coordinate system of all data points is constructed with the total amount of materials consumed as the horizontal axis and the progress of the project as the vertical axis; Preset the first neighborhood range of all data points; Determine the neighborhood anomaly degree of each data point according to the fluctuation degree of the first neighborhood range of each data point, and adjust the first neighborhood range of each data point to a second neighborhood range using the neighborhood anomaly degree of each data point; Evaluate the fitness of the second neighborhood range of each data point and calculate the third neighborhood range of each data point: , in the formula, For the The third neighborhood of data points, For the The degree of adaptation of the second neighborhood range of the data points, is the preset first neighborhood range, For the The second neighborhood range of data points; The robust weight of each data point is determined according to the third neighborhood range of each data point, and the least squares fitting is performed based on the robust weights of all data points to obtain the overall fitting curve. The material consumption and project progress of the current date are supervised and managed in real time according to the change in the slope of the overall fitting curve.

2. The coal construction project supervision and management method based on big data according to claim 1 is characterized in that: The method for determining the robust weight of each data point based on the third neighborhood range of each data point is: Perform curve fitting on all data points within the third neighborhood of each data point to obtain the local curve of each data point and calculate the corresponding local curve of the data point. The coordinate value and the original The difference in coordinate values, and the reciprocal of the absolute value of the difference is taken as the robust weight of the data point.

3. The coal construction project supervision and management method based on big data according to claim 1 is characterized in that: The degree of adaptation of the second neighborhood range of each data point satisfies the following relationship: , , For the The degree of adaptation of the second neighborhood of the data point and the total number of data points in the second neighborhood, , For the data points, The second neighborhood of the data point The local fitting error of the data points is , For the data points, The second neighborhood of the data point The degree of abnormality of the neighborhood of a data point, , For the The standard deviation of the local fitting error and the standard deviation of the neighborhood anomaly of all data points within the second neighborhood of the data point, , For the The mean of the local fitting errors and the mean of the neighborhood anomaly of all data points within the second neighborhood of a data point, is the natural exponential function, is the absolute value symbol.

4. The method for monitoring and managing coal construction projects based on big data according to claim 1 is characterized in that: The method for real-time supervision and management of the material consumption and engineering progress on the current date according to the slope change of the overall fitting curve is: The slope at each data point on the overall fitting curve is obtained, and all slopes are detected for anomalies using the box plot method; if the slope at the data point of the current date is abnormal, it is determined that the material consumption and project progress of the current date are abnormal, and an early warning notification is issued, and the consumed materials and project progress of the current date are checked and analyzed; if the slope at the data point of the current date is not abnormal, it is determined that the material consumption and project progress of the current date are not abnormal, and no early warning notification is issued.

5. The method for monitoring and managing coal construction projects based on big data according to claim 1 is characterized in that: The first neighborhood range, second neighborhood range, and third neighborhood range of each data point are all length values. Each data point is taken as the starting point, and the neighborhood range corresponding to the data point is obtained by intercepting forward according to the length value.

6. The method for monitoring and managing coal construction projects based on big data according to claim 3 is characterized in that: The degree of neighborhood anomaly of each data point is calculated based on the following formula: ； In the formula, For the The degree of abnormality of the neighborhood of a data point, For the The fluctuation degree of the first neighborhood range of the data point, For the The total number of data points within the first neighborhood of a data point, For the The first neighborhood of the data point The original data point Coordinate values, For the The first neighborhood of the data point The original data point Coordinate values, For the The original data points of all data points in the first neighborhood of the data point The mean of the coordinate values, is the preset first neighborhood range, is the absolute value symbol, is a natural exponential function.

7. The method for monitoring and managing coal construction projects based on big data according to claim 3 is characterized in that: The local fitting error is determined based on the following method: Perform curve fitting on all data points within the second neighborhood of each data point to obtain the local curve of each data point; calculate the corresponding local curve of the data point The coordinate value and the original The absolute value of the difference in coordinate values ​​is taken as the local fitting error of the data point.

8. The method for monitoring and managing coal construction projects based on big data according to claim 6 is characterized in that: The degree of fluctuation of the first neighborhood range of each data point is determined based on the following method: Calculate the first neighborhood of each data point. Variance of coordinate values , and all data points in the coordinate system Variance of coordinate values ;Will and The ratio of is taken as the fluctuation degree of the first neighborhood range of the data point.

9. The method for monitoring and managing coal construction projects based on big data according to claim 6, characterized in that: The method of adjusting the first neighborhood range of each data point to the second neighborhood range by using the neighborhood anomaly degree of each data point is: Preset neighborhood anomaly threshold ; Compute the second neighborhood of each data point: , where For the The second neighborhood of data points, is the preset first neighborhood range, For the The degree of abnormality of the neighborhood of a data point, The symbol for rounding up.

10. A coal construction project supervision and management system based on big data, characterized in that: The supervision management system includes a memory and a processor, the memory stores a computer program, and the processor executes the computer program to implement the steps of the supervision management method according to any one of claims 1 to 9.

Images