Visual analysis system for road construction information

Abstract

The invention relates to the technical field of data visualization, in particular to a road construction information visualization analysis system which comprises a construction state layering module, a construction mode analysis module, a construction trend prediction module, an error dynamic correction module and an information storage management module. According to the method, the real-time performance and interactivity of the construction state are improved by performing multi-dimensional dynamic analysis on the road construction information, and the information such as equipment operation, construction duration and state fluctuation is accurately identified and sorted by combining the hierarchical state model and construction characteristic weight value analysis; the future construction trend can be dynamically predicted and corrected in time, meanwhile, the construction information storage mode is optimized, data management is more accurate and reliable, the defects of static display and lack of real-time updating in a traditional scheme are overcome, the strain capacity and flexibility in a complex construction environment are enhanced, and the construction efficiency is improved. Continuous tracking and adjustment of the construction information are guaranteed, and the management precision and the decision timeliness are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data visualization, in particular to a road construction information visualization analysis system. BACKGROUND

[0002] The technical field of data visualization mainly involves presenting data in the form of graphics, images, animations, etc. to make it easier for users to understand and analyze data more intuitively and conveniently. The core of the technical field includes graphic design, data representation, user interaction, and visual expression of data analysis. Data visualization technology is applied to information display in various data processing processes, simplifying data content through graphical methods to make complex data more readable and operable. It is widely used in business, medical, transportation, environmental, social science, etc. Through effective visualization means, decision-makers and users can find rules and trends in massive data, improving the efficiency and accuracy of data analysis. Among them, the traditional road construction information visualization analysis system refers to a system for displaying and analyzing construction-related information during road construction through data visualization. Traditional road construction information visualization analysis presents construction progress, resource allocation, site management, etc. through static charts, simple maps, or two-dimensional view-based display methods. This traditional method lacks dynamic interactivity and real-time data updating capabilities, and cannot effectively respond to complex and changing construction environments and immediate decision-making needs.

[0003] In actual operation, the prior art relies on static charts and simple maps to display construction information. The traditional method lacks dynamic updating and real-time interactivity, and cannot respond to changing needs in complex construction environments. Construction progress, resource allocation, and site management information cannot be reflected in actual situations in a timely manner, resulting in inaccurate understanding of construction status during decision-making, especially when unexpected changes occur during construction. This limitation leads to low construction management efficiency, inability to adjust construction plans and resource scheduling in real time, and ultimately affects project progress and quality, making it impossible to achieve precise control and effective decision-making during construction. SUMMARY

[0004] The purpose of the present application is to solve the problems existing in the prior art and to provide a road construction information visualization analysis system.

[0005] To achieve the above purpose, the present application adopts the following technical solution: a road construction information visualization analysis system comprises:

[0006] The construction state layering module identifies node distribution characteristics of a construction area, divides construction state levels, extracts state change frequency and construction period distribution density, compares the state change frequency and the construction period distribution density, judges state attribution relationship, and obtains a layered construction state model based on multi-dimensional state data in a road construction process.

[0007] The construction mode analysis module calls the layered construction state model, extracts construction equipment operation data, construction time length data, and state fluctuation data under each classification, performs amplitude sorting on the equipment operation data and the state fluctuation data, filters front feature sorting, and obtains a construction characteristic weight value.

[0008] The construction trend prediction module arranges nodes in a future time period according to the construction characteristic weight value, extracts a construction growth trend of adjacent time periods, judges a trend offset angle, corresponds the offset angle and the time length, and obtains a reconstructed construction prediction trend table.

[0009] The error dynamic correction module extracts a prediction value and a real-time value difference interval in a time node based on the reconstructed construction prediction trend table, analyzes a difference interval direction consistency ratio and a difference amplitude ratio, judges a ratio of the two ratios, and obtains a construction information error dynamic correction data list.

[0010] As a further scheme of the present application, the layered construction state model includes a state density level, a construction period distribution weight, and a level label mapping relationship, the construction characteristic weight value includes an equipment operation fluctuation weight, a state response weight, and a construction frequency distribution weight, the reconstructed construction prediction trend table includes a construction growth direction sequence, a trend offset angle interval, and a time period change gradient, and the construction information error dynamic correction data list includes a prediction error adjustment parameter, a trend consistency evaluation value, and a dynamic correction ratio.

[0011] As a further scheme of the present application, the construction state layering module includes:

[0012] The construction period division sub-module divides a sudden change position and a state partition in a construction period based on multi-dimensional state data in a road construction process, including equipment start-stop records, construction logs, and state time node sequences, and obtains construction period division data.

[0013] The state frequency extraction sub-module filters construction logs in each partition according to the construction period division data, extracts a label frequency and identifies an average occurrence frequency, and obtains a state label frequency value.

[0014] The distribution density ratio calls the state label frequency value, analyzes the partition state node density and the state frequency, calculates a state and density matching difference value, compares and judges an attribution judgment value, identifies a construction period attribution division quantity, judges a partition attribution relationship, and obtains a layered construction state model.

[0015] As a further scheme of the present application, the construction mode analysis module comprises:

[0016] The construction data extraction submodule calls the layered construction state model, extracts equipment operation data, construction duration and state fluctuation data of each classification, performs field verification and converts into continuous variables, and obtains construction behavior characteristic quantities;

[0017] The sorting feature screening submodule compares the amplitude values of the equipment operation data and the state fluctuation data according to the construction behavior characteristic quantities, sorts the amplitude values in each classification, retains the feature indexes before sorting, and obtains high-frequency variation feature values;

[0018] The classification weight calculation submodule calls the construction duration data of the high-frequency variation feature values, extracts the cumulative values of the equipment operation and state fluctuation items before sorting, identifies the proportion of the cumulative values to the construction duration in each classification, summarizes the feature proportion items, and obtains construction characteristic weight values.

[0019] As a further scheme of the present application, the construction trend prediction module comprises:

[0020] The classification weight identification submodule extracts the construction flow direction in the node according to the construction characteristic weight values, identifies the flow increase and decrease interval values and the node category characteristics, analyzes the matching degree of the node and the growth trend, and obtains flow trend classification weight values;

[0021] The trend offset judgment submodule calls the flow trend classification weight values, extracts adjacent time period growth trend vectors, identifies the vector angle and the growth rate ratio, combines the time period duration, and obtains a trend offset angle matching degree;

[0022] The prediction trend analysis submodule extracts the construction trend and the direction angle in the matching degree interval according to the trend offset angle matching degree, reorganizes the extension in order, and obtains a reconstructed construction prediction trend table.

[0023] As a further scheme of the present application, the error dynamic correction module comprises:

[0024] The error interval extraction submodule identifies the difference between the real-time data and the predicted data of the node in the time period based on the reconstructed construction prediction trend table, calculates the error identification value of the node between the time periods, calibrates the node at the different time difference, divides the error interval and determines the upper and lower limit values, records the start and end time stamps of the error interval, extracts the error direction, and obtains an error direction interval sequence.

[0025] The consistency ratio calculation sub-module calculates the proportion of the absolute value of the error in the error direction duration interval according to the error direction interval sequence, and judges the error trend fluctuation to obtain an error trend deviation index;

[0026] The deviation data prediction sub-module calls the error trend deviation index, finds the error deviation index change area, fuses the corrected section and the original trend data, updates the construction trend curve, and obtains a construction information error dynamic correction data list.

[0027] As a further scheme of the present application, the system further comprises an information storage management module:

[0028] The information storage management module extracts the interval equipment identification field, the construction trajectory field and the timestamp field based on the construction information error dynamic correction data list, separates the fields according to the field type, disperses the equipment identification field to multiple nodes after strength encryption, records the encryption index path of the node corresponding to the encrypted data, and obtains a partition storage path of the construction analysis field;

[0029] The partition storage path of the construction analysis field comprises an encrypted equipment index path, a construction field storage location and a timestamp data partition mapping.

[0030] As a further scheme of the present application, the information storage management module comprises:

[0031] The field separation sub-module identifies the equipment record and analyzes the field content based on the construction information error dynamic correction data list, classifies and processes according to the field identification and labeling features, and generates a field classification and splitting result;

[0032] The distributed encryption sub-module calls the equipment identification field in the field classification and splitting result, extracts the sequence structure and field bitmap, performs strength encryption according to the mapping relationship, and disperses the encrypted data to multiple nodes, records the key path and node number of each encrypted field, and obtains encrypted node distribution data;

[0033] The path indexing sub-module calls the key path and node index in the encrypted node distribution data, classifies and maps the node path, arranges the index structure of the construction trajectory field and the timestamp field, and obtains a partition storage path of the construction analysis field.

[0034] Compared with the prior art, the present application has the following advantages and positive effects:

[0035] In the present application, by carrying out multi-dimensional dynamic analysis on road construction information, the real-time and interactivity of the construction state are effectively improved, the hierarchical construction state model and the analysis method of the construction characteristic weight value are adopted, so that the equipment operation data, construction time, state fluctuation and the like in the construction process are more accurately identified and sorted, and then dynamic prediction and timely correction of future construction trends are provided, the storage mode of construction information is improved, the data management is more accurate and reliable, the management efficiency of the construction process and the accuracy of the decision are greatly improved, the defects of static display and real-time update in the traditional scheme are solved in this way, the response capability and flexibility in the complex construction environment are significantly improved, the dynamic tracking and timely adjustment of the construction information are ensured, and thus the precision of the construction management and the timeliness of the decision are improved. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 is a system flowchart of the present application;

[0037] Figure 2 is a construction state hierarchical module flowchart in the present application;

[0038] Figure 3 is a construction mode analysis module flowchart in the present application;

[0039] Figure 4 is a construction trend prediction module flowchart in the present application;

[0040] Figure 5 is an error dynamic correction module flowchart in the present application;

[0041] Figure 6 is an information storage management module flowchart in the present application. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0043] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0044] Referring to Figure 1 A road construction information visual analysis system comprises:

[0045] The construction state layering module identifies node distribution characteristics of a construction area, divides construction state levels, extracts state change frequency and construction period distribution density between nodes and compares them, judges state attribution relationship, and obtains a layered construction state model based on multi-dimensional state data in a road construction process;

[0046] The construction mode analysis module calls the layered construction state model, extracts construction equipment operation data, construction duration data and state fluctuation data under each classification, sorts the equipment operation data and the state fluctuation data in amplitude, filters the front features of the sorted order, and obtains construction characteristic weight values;

[0047] The construction trend prediction module arranges nodes in a future time period according to the construction characteristic weight values, extracts construction growth trends of adjacent time periods and judges trend offset angles between them, corresponds the offset angles with time lengths, and obtains a reconstructed construction prediction trend table;

[0048] The error dynamic correction module extracts a prediction value and a real-time value difference interval in a time node based on the reconstructed construction prediction trend table, analyzes the direction consistency proportion and the difference amplitude proportion of the difference interval, judges the ratio of the two proportions, and obtains a construction information error dynamic correction data list;

[0049] The information storage management module extracts interval equipment identification fields, construction trajectory fields and time stamp fields based on the construction information error dynamic correction data list, separates the fields according to field types, disperses the equipment identification fields to multiple nodes after strength encryption, records the encryption index path of the corresponding encrypted data of the nodes, and obtains a partition storage path of the construction analysis field.

[0050] The layered construction state model comprises state density levels, construction period distribution weights, and level label mapping relationships. The construction characteristic weight values comprise equipment operation fluctuation weights, state response weights, and construction frequency distribution weights. The reconstructed construction prediction trend table comprises construction growth direction sequences, trend offset angle intervals, and time period change gradients. The construction information error dynamic correction data list comprises prediction error adjustment parameters, trend consistency evaluation values, and dynamic correction ratios. The partition storage path of the construction analysis field comprises encrypted equipment index paths, construction field storage locations, and time stamp data partition mapping.

[0051] Referring to Figure 2 The construction state layering module comprises:

[0052] The construction period division sub-module is based on multi-dimensional state data in the road construction process, including equipment start-stop records, construction logs, state time node sequences, mutation positions and state partitions in the construction period are delimited, and construction period division data is obtained;

[0053] Based on the multi-dimensional state data in the road construction process, including equipment start-stop records, construction logs, state time node sequences, first extract the original data, for example, for a typical asphalt road paving project, the equipment start-stop record contains the record that the paver P001 starts at 08:00:00 on July 1 and stops at 12:00:00, and the record that the roller R002 starts at 08:05:00 and stops at 11:30:00, the construction log records events such as "08:10:00 asphalt paving starts" or "12:30:00 lunch break", and the state time node sequence forms a list of events in chronological order, for example: P001 starts, R002 starts, asphalt paving starts, P001 stops, R002 stops, then the mutation positions and state partitions in the construction period are delimited, in order to identify the mutation positions, a sliding time window method based on event density is used, a sliding time window of, for example, 30 minutes is set, in the window, the total number of equipment state change events (such as start / stop) and key construction log entries (such as stage start / end) is counted, if the total number exceeds the preset "mutation threshold", the center time of the window is marked as a mutation position, the "mutation threshold" is set according to the data analysis of historical road construction projects, by analyzing the data of the past 100 similar projects, it is found that 5 or more equipment state changes or 2 or more key log events occur within 30 minutes, indicating a significant transition of the construction stage, the threshold is determined as 5 for equipment state change count and 2 for key log event count, if 6 equipment state changes and 1 key log event occur within a 30-minute window, the weighted count is 6+1x2=8, which exceeds the threshold 5, so the center point of the window is marked as a mutation position, for example, if the paver starts, the roller starts, the paver pauses, the roller pauses, the material truck arrives, the personnel rest, etc. 6 events are recorded within the window of 08:00:00 to 08:30:00, and there is a key log of "asphalt paving starts", the total number of events (assuming weighted) in the window exceeds the preset mutation threshold, then 08:15:00 (the midpoint of the window) is identified as a mutation position, after identifying all the mutation positions, the entire construction period is accurately divided into multiple "state partitions", each partition represents a relatively stable construction stage, for example, if 08:15:00 and 14:30:00 are identified as mutation positions, the whole day construction can be divided into 00:00:00-08:15:00 (such as preparation stage), 08:15:00-14:30:00 (such as core paving stage), 14:30:00-23:59:59 (such as end and maintenance stage), this process accurately obtains the construction period division data.

[0054] The state frequency extraction submodule extracts the label frequency and identifies the average occurrence frequency according to the construction period division data, and obtains the state label frequency value;

[0055] According to the construction period division data, the system processes each divided "state partition" in detail, for example, for the identified "core paving stage", all construction log entries occurring in this time period are accurately selected, this selection is completed by comparing the time stamp of the log entry with the start and end time stamp of the partition, for example, only the logs with time stamp greater than or equal to 08:15:00 and less than 14:30:00 are selected, in the selected logs, the frequency of the pre-defined "state label" is automatically identified and extracted, the state label is a keyword describing the construction activity, for example, "paving operation", "rolling operation", "material transportation", "equipment failure", "personnel rest", the label is counted, for example, in the construction log of the above-mentioned "core paving stage", if the "paving operation" label appears 4 times, the "rolling operation" label appears 2 times, the "material transportation" label appears 2 times, the "equipment temporary shutdown" label appears 1 time, and the "personnel rest" label appears 1 time, then the average occurrence frequency of each label is identified, which is calculated by dividing the occurrence frequency of each label by the total duration of the "state partition", for example, the average frequency of "paving operation" is 4 times / 375 minutes ≈ 0.0107 times / minute, the average frequency of "rolling operation" is 2 times / 375 minutes ≈ 0.0053 times / minute, this calculation process is repeated for all identified state labels, and finally the state label frequency value is obtained.

[0056] The distribution density comparison submodule calls the state label frequency value, analyzes the partition state node density and the state frequency, calculates the state and density matching difference value, compares the attribution judgment value and makes a judgment, identifies the construction period attribution division quantity, judges the partition attribution relationship, and obtains the hierarchical construction state model;

[0057] The state label frequency value is called, and the "state node density" and the corresponding "state frequency" of each partition are analyzed. The "node density" is obtained by calculating the sum of the equipment start-stop events and the key log events in the partition, and then dividing by the duration of the partition. For example, if a "core paving phase" lasts for 375 minutes, with 10 equipment start-stop events and 3 key log events, the node density of the partition is (10+3) / 375≈0.0347 events / minute. By comparing this density with the average state frequency of different construction phases and conducting joint analysis, the "state and density matching difference value" is calculated, which quantifies the consistency between the observed state frequency and the node density. The matching difference value is calculated by comparing the standardized state frequency vector of each partition with the node density vector. Specifically, the state frequency vector of each partition is compared with the frequency vector of the reference phase, and the node density is combined to finally calculate the matching difference value. This difference value reflects the degree of fit between the state label frequency and the node density. According to historical data analysis and expert experience, the matching difference value is compared with the preset "belonging judgment value" to determine which construction phase each partition belongs to. For example, if the matching difference value is below a certain threshold (such as 0.12), it is determined to be the "subgrade construction" phase; if it is between 0.12 and 0.25, it is the "asphalt paving" phase; and if it exceeds 0.25, it is the "abnormal or mixed" phase. Through this process, the construction phases can be accurately identified and divided, and finally a layered construction state model is formed.

[0058] Please refer to Figure 3 The construction mode analysis module includes:

[0059] The construction data extraction sub-module calls the layered construction state model to extract the equipment operation data, construction duration, and state fluctuation data of each category, performs field verification and converts them into continuous variables to obtain the construction behavior characteristic quantity;

[0060] A layered construction state model is called, which defines each phase in the construction project in detail, for example, identifying the entire "asphalt paving" phase from July 1 to July 7, based on which the specific construction data under each category is accurately extracted, including "equipment operation data" (such as real-time sensor data of engine speed, fuel consumption rate, travel speed, and vibration frequency of the paver and roller), "construction duration" (i.e., the exact duration of each category phase, for example, the "asphalt paving" phase lasted for 6 days and 150 hours), and "state fluctuation data" (for example, the fluctuation of quality control parameters such as real-time temperature of asphalt mixture, paving thickness, and compaction degree during paving), for example, for the "asphalt paving" phase, the speed (RPM) records of all equipment P001 and the asphalt temperature data recorded by the infrared sensor installed on the paving board are pulled from the data bus, then field verification is performed, and converted into continuous variables, which involves integrity checks on the extracted data, such as ensuring that all time-stamped data points exist, and identifying and correcting or marking abnormal values, for example, the asphalt temperature sensor records a non-physical temperature of 500°C at a certain moment, which is marked as invalid or corrected to a reasonable boundary value (e.g., 180°C), after data cleaning, all discrete or categorical data is converted into continuous variables, for example, if the equipment operation state is originally a discrete state of "running", "idle", and "stopped", it is converted into a continuous variable of equipment power consumption percentage, or the graded compaction degree data (such as 1-5 levels) is converted into a specific compaction degree value percentage, ensuring that all feature quantities used for subsequent analysis are quantifiable and continuous, and the construction behavior feature quantity is obtained.

[0061] The sorting feature screening submodule compares the amplitude values of the equipment operation data and the state fluctuation data based on the construction behavior feature quantity, sorts the amplitude values within each category, retains the feature indicators before sorting, and uses the formula:

[0062] ;

[0063] The high-frequency fluctuation feature value is obtained;

[0064] wherein, represents the high-frequency fluctuation feature value, represents the amplitude value of the equipment operation data of the i-th data point, represents the mean value of the equipment operation data, represents the standard deviation of the equipment operation data, represents the amplitude value of the state fluctuation data of the i-th data point, represents the mean value of the state fluctuation data, represents the standard deviation of the state fluctuation data, ​​Total number of data points;

[0065] According to the construction behavior characteristic quantity, wherein the "amplitude value" specifically refers to the instantaneous measurement value of the equipment operation parameter or state fluctuation parameter at a certain moment, for example, the travel speed (m / min) of the paver at a certain second or the real-time temperature (℃) of the asphalt mixture at a certain position, by transverse comparison of the instantaneous amplitude value to identify the relative size in the overall data distribution, then, sort the amplitude value within each category, this sorting is not simply ascending or descending, but weighted sorting based on its frequency and importance in a specific construction phase, for example, in the "asphalt paving" phase, if the paver speed runs in the range of 4.0-4.5 m / min for 80% of the total running time, and this speed interval is closely related to the optimal paving efficiency, then this speed interval is determined as a high importance feature and is kept in the front position of the sorting, the feature indicators before the reserved sorting are the key parameters and their typical value ranges representing the core operation mode and state characteristics of the construction phase, then the formula Calculate the high-frequency fluctuation characteristic value, wherein, represents the high-frequency fluctuation characteristic value, which quantifies the product influence of the instantaneous deviation of equipment operation and the overall volatility of construction state, represents the amplitude value of the equipment operation data of the th data point, for example, the instantaneous speed (m / min) of the paver at a certain moment, represents the mean value of the equipment operation data, for example, the average speed of the paver throughout the "asphalt paving" phase, represents the standard deviation of the equipment operation data, which measures the dispersion of the equipment operation data relative to the mean value, reflecting the stability of the equipment operation, represents the amplitude value of the state fluctuation data of the th data point, for example, the instantaneous temperature (℃) of the asphalt mixture at a certain moment, represents the mean value of the state fluctuation data, for example, the average temperature of the asphalt throughout the phase, represents the standard deviation of the state fluctuation data, which measures the dispersion of the state fluctuation data relative to the mean value, reflecting the stability of the construction quality or environmental parameters, represents the total number of data points, i.e. the total number of state fluctuation data points collected in the current construction phase;

[0066] The operation logic of the formula is: first, part calculates the standardized absolute deviation of a single equipment operation data point relative to its mean value , i.e. the degree of deviation of the point from the average level, and is normalized by the standard deviation so that equipment data of different dimensions can be compared, secondly, Part of the calculation of a single state fluctuation data point Relative to its average value The standardized absolute deviation, and by standard deviation Normalization, again, The sum of the squares of all standardized state fluctuation deviations, which makes all deviations contribute in positive values, and emphasizes larger fluctuations, and finally, by taking the square root, the sum is returned to the original dimension, and multiplied by the standardized deviation of the device operation to obtain the comprehensive high-frequency variation characteristic value

[0067] The benefit of this formula is that it combines the instantaneous variation of the device operation with the overall volatility of the construction state, through the form of product, it can highlight those key moments when the device operation appears significant instantaneous deviation and at the same time accompanied by larger fluctuations in the construction state, thus more comprehensively reflecting the dynamic changes and potential problems in the construction process, rather than only focusing on a single dimension, for example, for a 1-hour asphalt paving operation, the total number of data points , the specific data is shown in Table 1;

[0068] First, calculate the mean value of the paver speed Meters per minute;

[0069] Its standard deviation Meters per minute;

[0070] Next, calculate the mean value of the asphalt temperature Celsius;

[0071] Its standard deviation Celsius;

[0072] Then, calculate the square sum of the state fluctuation term , for example, , sum all 10 data points to get , and finally, calculate , taking the first data point ( , ) as an example , the calculated , , , and values are brought into the formula:

[0073] This calculation process is carried out for each data point, and finally the high-frequency variation characteristic value is obtained;

[0074] Table 1: Example table of asphalt paving equipment operation and state fluctuation data: ​​

[0075]

[0076] As shown in Table 1, the data will be used to calculate the high-frequency fluctuation characteristic value.

[0077] The classification weight calculation submodule calls the construction duration data of the high-frequency fluctuation characteristic value, extracts the cumulative value of the equipment operation and state fluctuation item before sorting, identifies the proportion of the cumulative value in each classification and the construction duration, summarizes the characteristic proportion item, and obtains the construction characteristic weight value;

[0078] The construction duration data of the high-frequency fluctuation characteristic value is called, and the "pre-sorting characteristic index" is the key operation mode and state characteristic identified by the previous module, for example, "paver speed in the 4.0-4.5 m / min interval" and "asphalt temperature in the 150-160℃ interval". The cumulative duration of the characteristic in the entire "asphalt paving" stage (for example, the total construction duration is 375 minutes) is accurately calculated, for example, if the paver runs in the speed interval of 4.0-4.5 m / min for 300 minutes, and the asphalt temperature in the interval of 150-160℃ lasts for 320 minutes, then the proportion of the cumulative value in each classification and the construction duration is identified. This proportion quantifies the relative importance or universality of each characteristic in the entire construction stage, for example, the proportion of the paver speed in the 4.0-4.5 m / min interval is 300 minutes / 375 minutes=0.80, and the proportion of the asphalt temperature in the 150-160℃ interval is 320 minutes / 375 minutes≈0.853. The characteristic proportion item is summarized, and this summary process can use weighted average or aggregation method, for example, if the weights of the paver speed and the asphalt temperature are equal, then the comprehensive weight value is (0.80+0.853) / 2=0.8265, which reflects the overall compliance or stability of the construction stage in terms of key operation parameters and quality control. Finally, the construction characteristic weight value is obtained.

[0079] Please refer to Figure 4 , the construction trend prediction module includes:

[0080] The classification weight identification submodule extracts the construction flow direction in the node according to the construction characteristic weight value, identifies the flow increase and decrease interval value and the node classification characteristic, analyzes the matching degree of the node and the growth trend, and obtains the flow trend classification weight value.

[0081] According to the construction characteristic weight value, the weight value reflects the typical operation mode and state characteristics of each construction stage. First, the construction flow direction in the node is extracted, where the "node" refers to the key milestone or time point on the project timeline, such as the completion of a certain length of road section or the daily closing point, and the "construction flow direction" quantifies the progress of the project between nodes, such as the daily paving length (m / day) or the total amount of materials consumed (tons / day). Then, the flow increase and decrease interval value is identified and the node category characteristics are identified. The flow increase and decrease interval value is determined by comparing the construction flow direction change between adjacent nodes, for example, if 150 meters are paved the previous day and 180 meters are paved the next day, the flow increase is 30 meters. The node category characteristics classify each node, such as "daily completion node", "stage acceptance node", "equipment failure node", etc. The characteristics provide additional context information about the nature of the node. Then, the matching degree of the node and the growth trend is analyzed, which involves comparing the current "flow increase and decrease interval value" with the predefined "growth trend matching threshold value", for example, the daily paving length increase is matched with the "normal growth trend" (defined as an increase range of 150-200 meters per day). This "growth trend matching threshold value" is set based on historical project data and industry standards, for example, by analyzing the daily progress data of the past 100 projects, it is found that when the daily paving length increase is within ±10% of the planned value, the project progress is considered normal, therefore, if the planned daily increase is 175 meters, the matching threshold range is set to [157.5, 192.5] meters. For example, if the paving length increase of a "daily completion node" is 165 meters, its matching degree with the "normal growth trend" is high, while if the increase is 0 meters and the node category is "equipment failure node", its matching degree with any positive growth trend is very low. This analysis process accurately obtains the flow trend classification weight value.

[0082] The trend deviation judgment submodule calls the flow trend classification weight value, extracts the adjacent time period growth trend vector, identifies the vector angle and growth rate ratio, and combines the time period duration to obtain the trend deviation angle matching degree.

[0083] In the call of the traffic trend classification weight value, the construction progress of each day is represented by a vector containing the completed work and resource consumption, for example, the first day is to complete 150 meters and consume 500 liters of oil, and the second day is to complete 180 meters and consume 550 liters of oil. Then, the included angle of the vector is compared with the growth rate ratio: the vector included angle reflects the consistency of the construction direction, the smaller the included angle, the more stable the trend, and the larger the included angle, the more significant the direction change; the growth rate ratio is used to measure the degree of progress acceleration or deceleration, for example, from 150 meters to 180 meters means about 20% acceleration. On this basis, the duration of the time period is also weighted to highlight the importance of long-term trends, for example, a slight acceleration trend lasting 5 days is more meaningful than a significant acceleration trend lasting only 1 day, and in this way, the matching degree of the trend deviation angle can be calculated by considering the construction direction, growth speed and duration, so as to accurately reflect the matching degree of the construction process and the predetermined growth trend, and obtain the trend deviation angle matching degree.

[0084] The prediction trend analysis submodule extracts the construction trend and direction angle of the matching degree interval according to the trend deviation angle matching degree, reorganizes and extends them in order to obtain the reconstructed construction prediction trend table;

[0085] According to the trend deviation angle matching degree, the matching degree quantifies the deviation between the current construction trend and the expected trend. First, the construction trend and the direction angle of the matching degree interval are extracted. Specifically, the calculated "trend deviation angle matching degree" value is mapped to the predefined "matching degree interval". Each interval is associated with a specific "construction trend" (such as "stable growth", "slight fluctuation", "significant deviation") and a "direction angle". This "direction angle" represents the angle of the actual trend deviating from the ideal path. For example, if the matching degree is between 0.0 and 0.1, it is identified as a "stable growth" trend with a direction angle of 0 degrees. If the matching degree is between 0.1 and 0.3, it is identified as a "slight fluctuation" trend with a direction angle of 5 degrees. If the matching degree is greater than 0.5, it is identified as a "significant deviation" trend with a direction angle of more than 20 degrees. The interval and the corresponding trend and angle are set based on statistical analysis of the impact of different deviation degrees on the final project in historical construction data. Subsequently, the sequence is reorganized and extended. This means that the "construction trend" and "direction angle" identified are used to sequentially adjust and extrapolate the future construction progress in the "reconstructed construction prediction trend table". For example, if the matching degree of the current and recent time periods indicates "slight fluctuation" and the direction angle is stable at 5 degrees, the ideal linear growth will not be assumed when predicting future construction trends. Instead, this 5-degree slight deviation trend will be incorporated into the future prediction curve. Through sequential analysis of historical trends and direction angles, patterns can be identified and used to more accurately extrapolate future trends. For example, if it is observed that the daily completion length growth trend for three consecutive days is 150 meters, 160 meters, and 155 meters, respectively, and the corresponding "trend deviation angle matching degree" is in the "slight fluctuation" interval, it is predicted that the next day will also be within this fluctuation range, and the expected completion length will be adjusted accordingly. This adjustment ensures that the prediction result is more consistent with the actual operating conditions, and finally the reconstructed construction prediction trend table is obtained.

[0086] Please refer to Figure 5 , the error dynamic correction module includes:

[0087] The error interval extraction submodule identifies the difference between the real-time data and the predicted data of the node corresponding to the time period based on the reconstructed construction prediction trend table. The formula is:

[0088] ;

[0089] The error identification value of the node between time periods is calculated, the time difference when the node is differentiated is calibrated, the error interval is divided and the upper and lower limit values are determined, the start and end time stamps of the error interval are recorded, the error direction is extracted, and the error direction interval sequence is obtained;

[0090] wherein, represents the error identification value of the node between time periods, represents the node real-time data of the node, representative node predicted data of the node, average value of real-time data of all nodes in the time period, duration of the time period, a constant factor, maximum value of duration of all time periods, total number of nodes;

[0091] Based on the reconstructed construction prediction trend table, which provides updated construction progress prediction data, first identify the difference between the real-time data and the predicted data of the node corresponding to the time period. Here, "node" refers to a specific time point in the prediction trend table, such as the amount of construction completed every hour or every four hours. Accurately compare the "real-time data" ( ) actually recorded at each node with the "predicted data" ( ), for example, in a day (24 hours) of construction, if it is predicted that 12 meters will be completed in an hour, but the real-time data is only 10 meters, there is a difference. Then use the formula to calculate the error identification value of the node between time periods, where, error identification value of the node between time periods, which quantifies the standardized size of the prediction bias and its time persistence effect, real-time data of the node, for example, the actual construction length (unit: meters) completed at node , predicted data of the node, for example, the predicted construction length (unit: meters) to be completed at node , average value of real-time data of all nodes in the current time period, for example, the average value of the length completed in each hour in a 24-hour time period, which provides a basis for subsequent standardized error calculation, duration of the current time period, for example, the time period for the current error calculation is 24 hours, a constant factor, its value is 0.5, which is determined based on sensitivity analysis of multiple historical construction cases, aiming to ensure that systematic errors over a long period of time are fully weighted, while avoiding overreaction to transient fluctuations, maximum value of duration of all considered time periods, for example, the longest continuous prediction period duration found in historical data is 72 hours, used to normalize the duration of the current time period, total number of all nodes used for calculation in the current time period, for example, if data is collected every hour in 24 hours, ;

[0092] The operational logic of this formula is as follows: the first part of the formula Calculated the nodes The standardized value of the absolute difference between real-time data and predicted data, where the denominator is... This represents the overall volatility of real-time data within the current time period, making the error values ​​comparable across construction projects of different scales. The second part of the formula... It is a time-weighted factor, which varies with the duration of the error. It increases with the growth, when The closer The greater the weight given to error correction, the more this design ensures that persistent systematic errors receive more significant attention and correction than sporadic errors.

[0093] The advantage of this formula lies in the fact that it not only quantifies the magnitude of instantaneous prediction errors, but also, by introducing a time persistence factor, enables the model to distinguish and prioritize the correction of errors that have accumulated over a long period of time and exhibit systematic deviation characteristics. This significantly improves the long-term stability and correction effect of the prediction model. For example, assuming that there are data for five key nodes during a 24-hour construction period as shown in Table 2, the average value of the real-time data is first calculated. rice;

[0094] Next, calculate the denominator. ;

[0095] Then calculate for each node. , with nodes For example, its real-time data Predicted data Duration of time period Hours, maximum duration Hours, constant factor ;

[0096] Substituting into the formula, we get: ;

[0097] Similarly, node 2 Node 3 Node 4 Node 5 Then, the node time difference at the differentiated time is calibrated, and the error interval is divided to determine the upper and lower limits. Based on the calculated... The value divides the error into different intervals, for example, by... Determined to be in the "low error range", Determined as "mean error range", The threshold is set according to the statistical analysis result of the historical project error tolerance, for example, through analysis of the past project data, it is found that when The value is less than 0.06, the final error rate of the project is controlled within 2%, and when The value is higher than 0.11, the error rate of the project may exceed 5%, for example, in the above calculation result, Belongs to "low error interval", Belongs to "medium error interval", Belongs to "high error interval", record the start and end time stamp of the error interval, for example, record the "high error interval" from the current node to the next stable interval, at the same time, extract the error direction, that is, whether the real-time data is higher or lower than the predicted data, for example, if It is positive error, if It is negative error, this process accurately obtains the error direction interval sequence;

[0098] Table 2: Construction prediction and real-time data comparison example table:

[0099]

[0100] As shown in Table 2, this is the node real-time data and predicted data used for error calculation.

[0101] The consistency ratio calculation submodule calculates the duration of the direction interval and the direction switching density according to the error direction interval sequence, calculates the proportion of the absolute value of the error in the error direction continuous interval, judges the error trend fluctuation, and obtains the error trend deviation index;

[0102] According to the error direction interval sequence, the sequence contains the amplitude, direction and duration information of the error, first, the duration of the direction interval and the direction switching density are counted, for example, in a 24-hour analysis period, if the "negative error" lasts for 4 hours, then the "positive error" lasts for 6 hours, and then changes to "negative error" for 3 hours, the duration of each direction interval is recorded, and the direction switching density is calculated, for example, in 24 hours, there are 2 times of direction switching, then the density is 2 times / 24 hours≈0.083 times / hour, then the proportion of error absolute value in the error direction duration interval is calculated, this calculation is obtained by dividing the sum of error absolute value of all nodes in each duration interval by the total sum of error absolute value in the entire analysis period, for example, if the total sum of error absolute value in the "negative error" 4-hour interval is 5 meters, and the total error absolute value in 24 hours is 20 meters, then the proportion of this interval is 5 meters / 20 meters=0.25, then the error trend fluctuation is judged, which is based on "direction switching density" and "proportion of error absolute value", for example, if the "direction switching density" exceeds the preset "high volatility threshold" (such as 0.1 times / hour), and the proportion of single direction error does not reach the "dominance threshold" (such as 0.6), it is determined that the error trend is "high volatility", this "high volatility threshold" and "dominance threshold" are empirically set according to the influence of error mode on prediction stability in historical project data, for example, by analyzing a large amount of historical error data, it is found that when the direction switching density exceeds 0.1 times / hour, the prediction accuracy will be significantly reduced, indicating that the error mode tends to be random fluctuation, if the "direction switching density" of a period is 0.125 times / hour (exceeding the threshold of 0.1), and the maximum error absolute value proportion is 0.385 (not reaching the dominance threshold of 0.6), it is determined that the error trend of this period has "high volatility", and finally the error trend deviation index is obtained.

[0103] The offset data prediction sub-module calls the error trend deviation index, finds the error deviation index change area, fuses the corrected section and the original trend data, updates the construction trend curve, and obtains the construction information error dynamic correction data list;

[0104] The error trend offset index is called, which quantitatively describes the stability and systematicness of the prediction error. First, find the error offset index change area, which involves continuous monitoring of the dynamic changes of the "error trend offset index", and identifying the specific time period when the value of the index changes significantly, for example, the index suddenly jumps from a long-term stable low value (such as 0.05) to a high value (such as 0.15), exceeding the preset "change area threshold", then this area is marked as "change area", this "change area threshold" is set according to the analysis of the fluctuation characteristics of the index in the actual project, for example, when the rolling average of the "error trend offset index" in the past 24 hours changes by more than 10%, it is considered to enter the "change area", this 10% change amplitude is determined according to the minimum significant change in historical data that can indicate that the future prediction model needs to be adjusted, then fuse the corrected section with the original trend data, update the construction trend curve, and generate a correction factor in the identified "change area" or based on the latest "error trend offset index" to correct the future prediction, for example, if the "error trend offset index" shows that the prediction is consistently 5% overestimated, then all subsequent construction quantities will be multiplied by 0.95 to be adjusted down, this correction factor is directly related to the "error trend offset index", this process accurately obtains the construction information error dynamic correction data list.

[0105] The "construction information error dynamic correction data list" is used to monitor and correct information errors that occur during construction. By dynamically collecting and analyzing data errors during construction, it helps identify potential problem areas and generates correction plans. The list provides detailed data on various construction errors, such as position deviation and time error, and adjusts construction strategies in a timely manner to ensure the improvement of construction precision and quality. Through continuous tracking and correction of errors, it can effectively reduce the deviation that may occur during construction, thereby optimizing the overall construction process and results.

[0106] Please refer to Figure 6 , the information storage management module includes:

[0107] The field separation sub-module identifies device records and parses field content based on the construction information error dynamic correction data list, classifies and processes them according to field identification and labeling features, and generates field classification and splitting results;

[0108] Based on the construction information error dynamic correction data list, which contains the dynamically corrected construction prediction and real-time data information, first identify the equipment records and parse the field content, which involves identifying the data entries generated by specific equipment from the original data stream, for example, identifying all data records belonging to paver P001 or roller R002, and then parsing the content of the records, breaking down semi-structured or unstructured raw data such as "P001, 07-1509:30:00, Speed=4.2m / min, Temp=155C" into independent and identifiable fields such as "device ID", "timestamp", "speed", "temperature", etc., and classifying them according to field identifiers and labeling features, where "field identifier" is a predefined field name such as "Speed", and "labeling feature" is metadata related to the field, such as data type (numeric, string), unit of measurement (m / min, °C), data sensitivity (high, medium, low), etc. According to the identifier and feature, the field is classified, for example, "device ID" is classified as "identification data", "timestamp" is classified as "time data", and "speed" and "temperature" are classified as "operation parameter data", and according to their impact on project progress, they are marked as "criticality: high" or "criticality: medium". This process accurately generates field classification and splitting results.

[0109] The distributed encryption sub-module calls the device identifier field in the field classification and splitting result, extracts the sequence structure and field bitmap, performs strength encryption according to the mapping relationship, and disperses the encrypted data to multiple nodes, records the key path and node number of each encrypted field, and obtains encrypted node distribution data.

[0110] By analyzing the device identifier field (such as "P001"), its composition structure is extracted, "P" is identified as the device type and "001" is identified as the serial number, and a field bitmap is generated to mark sensitive and non-sensitive information. The sensitive part (such as "001") will be encrypted, and the non-sensitive part (such as "P") will remain unchanged. Encryption uses a predefined mapping relationship that associates field type, encryption algorithm (such as AES-256), and key management rules. Specifically, for the sensitive part of "P001", use AES-256 algorithm and preset key (such as K_P001) to encrypt and generate ciphertext. The encrypted data is dispersed to multiple nodes according to the distributed storage strategy, for example, the first half of the ciphertext is stored in node A and the second half is stored in node B, while recording the decryption key path and node number of each piece of data to ensure the accuracy of the subsequent secure reconstruction and decryption process. Through this process, efficient and secure device identifier encryption and storage are achieved, and encrypted node distribution data is obtained.

[0111] The path indexing sub-module calls the key path and node index in the encrypted node distribution data, classifies and maps the node path, arranges the index structure of the construction trajectory field and the timestamp field, and obtains the partition storage path of the construction analysis field.

[0112] The key path and node index in the encrypted node distribution data are called, which accurately describes the storage location of the encrypted data segment and the access path of the decryption key. First, the node path is classified and mapped. Here, the "node path" refers to the address of the physical or logical storage location where the encrypted data or key is stored. According to the predefined rules (such as data sensitivity, device type, geographic location, etc.), the path is classified, for example, all paths storing "core confidential device ID" data are classified as "high security area path", and the path is mapped to a specific security access strategy and storage partition. Then, the index structure of the construction trajectory field and the timestamp field is arranged. This operation is aimed at non-encrypted or independently encrypted "construction trajectory field" (such as device GPS coordinates: longitude, latitude, altitude) and "timestamp field". In order to construct an efficient index structure, for example, an R-tree spatial index is constructed for geographic coordinate data to quickly retrieve all construction activities in a certain geographic area, and a B-tree time index is constructed for timestamp data to efficiently query construction records in a certain time period. For example, in order to quickly query the work trajectory of all pavers in a certain road section during July, the system will construct a composite index combining timestamp and geographic spatial information. This index structure will ensure that even if sensitive information is encrypted and stored in scattered storage, non-sensitive but crucial trajectory and time information for analysis can still be efficiently retrieved and utilized. Finally, the partition storage path of the construction analysis field is obtained.

[0113] The above is only a preferred embodiment of the present application, and does not limit the form of the present application. Any skilled person in the art can use the disclosed technical content to make changes or modifications to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made according to the technical essence of the present application to the above embodiments without departing from the technical solution content of the present application still belongs to the protection scope of the present application.

Claims

1. A road construction information visual analysis system, characterized by, The system comprises: The construction state layering module identifies node distribution characteristics of a construction area, divides construction state levels, extracts state change frequency and construction cycle distribution density, and compares them to determine state attribution relationship based on multi-dimensional state data in a road construction process, to obtain a layered construction state model; The construction mode analysis module calls the layered construction state model, extracts construction equipment operation data, construction duration data, and state fluctuation data under each classification, sorts the equipment operation data and state fluctuation data by amplitude, filters the top features, and obtains construction characteristic weight values; The construction trend prediction module arranges construction flow in nodes in future time periods according to the construction characteristic weight values, extracts construction growth trends of adjacent time periods and judges the trend offset angle, and corresponds the offset angle with the time length to obtain a reconstructed construction prediction trend table; The error dynamic correction module extracts the prediction value and real-time value difference interval in the time node based on the reconstructed construction prediction trend table, analyzes the difference interval direction consistency proportion and the difference amplitude proportion, judges the ratio of the two proportions, and obtains a construction information error dynamic correction data list.

2. The road work information visual analysis system of claim 1, wherein, The layered construction state model comprises state density level, construction cycle distribution weight, and level label mapping relationship, the construction characteristic weight values comprise equipment operation fluctuation weight, state response weight, and construction frequency distribution weight, the reconstructed construction prediction trend table comprises construction growth direction sequence, trend offset angle interval, and time period change gradient, and the construction information error dynamic correction data list comprises prediction error adjustment parameter, trend consistency evaluation value, and dynamic correction ratio.

3. The road work information visual analysis system of claim 1, wherein, The construction state layering module comprises: The construction cycle division sub-module divides the construction cycle based on multi-dimensional state data in a road construction process, including equipment start-stop records, construction logs, and state time node sequence, and determines mutation positions and state partitions in the construction cycle to obtain construction cycle division data; The state frequency extraction sub-module filters the construction logs in each partition according to the construction cycle division data, extracts label frequency and identifies average occurrence frequency to obtain state label frequency value; The distribution density comparison sub-module calls the state label frequency value, analyzes partition state node density and state frequency, calculates state and density matching difference value, compares attribution judgment value and judges, identifies construction cycle attribution division quantity, judges partition attribution relationship, and obtains a layered construction state model.

4. The road work information visual analysis system of claim 3, wherein, The construction mode analysis module comprises: The construction data extraction sub-module calls the layered construction state model, extracts equipment operation data, construction duration, and state fluctuation data of each classification, performs field verification and converts them into continuous variables to obtain construction behavior characteristic quantity; The sorting feature filtering sub-module compares the amplitude values of the equipment operation data and state fluctuation data according to the construction behavior characteristic quantity, sorts the amplitude values in each classification, retains the feature indexes before sorting, and obtains high-frequency variation characteristic values; The classification weight calculation submodule calls the construction duration data of the high-frequency fluctuation characteristic value, extracts the cumulative value of the equipment operation and state fluctuation item before sorting, identifies the proportion of the cumulative value in each classification and the construction duration, summarizes the characteristic proportion item, and obtains the construction characteristic weight value.

5. The road work information visual analysis system of claim 4, wherein, The construction trend prediction module comprises: The classification weight identification submodule extracts the construction flow direction in the node according to the construction characteristic weight value, identifies the flow increase and decrease interval value and the node category characteristic, analyzes the matching degree of the node and the growth trend, and obtains the flow trend classification weight value; The trend offset judgment submodule calls the flow trend classification weight value, extracts the adjacent time period growth trend vector, identifies the vector angle and the growth rate ratio, combines the time period duration, and obtains the trend offset angle matching degree; The prediction trend analysis submodule extracts the construction trend and the direction angle in the matching degree interval according to the trend offset angle matching degree, reorganizes the extension in order, and obtains the reconstructed construction prediction trend table.

6. The road work information visual analysis system of claim 5, wherein, The error dynamic correction module comprises: The error interval extraction submodule identifies the difference between the real-time data and the predicted data of the node in the time period, calculates the error identification value of the node between the time periods, calibrates the time difference of the node in the difference, divides the error interval and determines the upper and lower limit values, records the start and end time stamps of the error interval, extracts the error direction, and obtains the error direction interval sequence based on the reconstructed construction prediction trend table; The consistency ratio calculation submodule calculates the proportion of the absolute value of the error in the error direction continuous interval, judges the error trend fluctuation, and obtains the error trend offset index according to the error direction interval sequence, the duration of the direction interval and the direction switching density, and the error direction continuous interval. The offset data prediction submodule calls the error trend offset index, finds the error offset index change area, fuses the corrected section and the original trend data, updates the construction trend curve, and obtains the construction information error dynamic correction data list.

7. The road work information visual analysis system of claim 1, wherein, The system further comprises an information storage management module: The information storage management module extracts the interval equipment identification field, the construction trajectory field and the time stamp field based on the construction information error dynamic correction data list, separates the fields according to the field types, disperses the equipment identification field strength to multiple nodes after encryption, records the encryption index path of the corresponding encrypted data of the node, and obtains the partition storage path of the construction analysis field; The partition storage path of the construction analysis field comprises an encrypted equipment index path, a construction field storage location and a time stamp data partition mapping.

8. The road work information visual analysis system of claim 7, wherein, The information storage management module comprises: The field separation submodule identifies the equipment record and analyzes the field content based on the construction information error dynamic correction data list, classifies and processes the field content according to the field identification and labeled features, and generates a field classification and splitting result; The distributed encryption submodule calls the equipment identification field in the field classification and splitting result, extracts the sequence structure and field bitmap, performs strength encryption according to the mapping relationship, disperses the encrypted data to multiple nodes, records the key path and node number of each encrypted field, and obtains the encrypted node distribution data; The path indexing sub-module calls the key path and node index in the encryption node distribution data, classifies and maps the node path, arranges the construction track field and the index structure of the timestamp field, and obtains the partition storage path of the construction analysis field.

Images