A construction hoist intelligent control method and system
Through intelligent control methods, the position and scheduling of the resident layer of the construction elevator is optimized, and the problems of low efficiency, large energy consumption and poor safety of the traditional construction elevator control system are solved, and more efficient and safe construction operations are achieved.
Patent Information
- Application Number
- CN202411237833.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-05
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-09-05
AI Technical Summary
The traditional construction elevator control system relies on manual operation, has low operating efficiency, high energy consumption, poor safety, and cannot dynamically adjust the operating strategy according to actual construction needs, resulting in unreasonable resource allocation and reducing construction efficiency.
By obtaining elevator usage data and construction period data, feature extraction, graph construction, resident layer calculation and optimization are performed, resident layer location and scheduling algorithm are optimized, waiting time and energy consumption are reduced, construction efficiency and safety are improved sex.
Through intelligent algorithms, optimize the residency layer position and scheduling of the elevator, reduce wait time and running time, reduce energy consumption, improve construction efficiency and safety, extend equipment life, and reduce labor costs and failure rates.
Smart Images

Figure CN118753941B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent control technology, and in particular to an intelligent control method and system for a construction elevator. Background Art
[0002] With the development of the modern construction industry, the demand for intelligent mechanical equipment on construction sites is increasing, especially for construction hoists, which are indispensable vertical transportation equipment on construction sites. The optimization of their intelligent control systems is particularly important. At present, the operation of traditional construction hoists relies on manual control, which has many shortcomings, such as low operating efficiency, high energy consumption, and poor safety.
[0003] Traditional construction elevator control systems mainly rely on manual operation and simple automated control methods, lacking in-depth analysis and intelligent optimization of usage data. For example, during peak usage periods, due to the lack of real-time monitoring and effective management of elevator load conditions and usage frequency, problems such as long waiting times and excessive wear of equipment often occur. In addition, traditional systems cannot dynamically adjust the elevator's operating strategy according to actual construction needs, resulting in irrational resource allocation and reduced construction efficiency. Summary of the invention
[0004] In order to solve the above technical problems, the present invention proposes an intelligent control method and system for a construction elevator to solve at least one of the above technical problems.
[0005] The present application provides a construction elevator intelligent control method, comprising the following steps:
[0006] Step S1: Obtaining elevator usage data and construction period data;
[0007] Step S2: extracting usage features based on the elevator usage data to obtain elevator usage feature data, and constructing a construction elevator usage graph based on the elevator usage feature data and the construction period data to obtain construction elevator usage graph data;
[0008] Step S3: Calculate the resident layer according to the construction elevator usage map data to obtain the construction elevator resident map data; obtain the current construction period data, and match the current construction period data and the construction elevator resident map data to obtain the construction elevator resident recommendation data;
[0009] Step S4: Acquire current elevator usage data, perform residence optimization based on the current elevator usage data and construction elevator residence recommendation data, and obtain current construction elevator residence optimization data to perform intelligent control operations on the construction elevator.
[0010] In the present invention, the resident layer position of the elevator is optimized through an intelligent algorithm, which reduces the waiting time and running time of the elevator, thereby improving the overall construction efficiency. The optimized resident layer position and intelligent scheduling algorithm reduce unnecessary operation of the elevator, reduce energy consumption, and save energy. Real-time monitoring and data analysis can detect abnormal situations in time, prevent potential failures and safety hazards, and improve the safety of the construction site. Dynamic adjustment is carried out in combination with real-time data and construction period data, so that the system can flexibly respond to changes in construction progress and emergencies and maintain efficient operation. Intelligent control operations reduce dependence on manual operation, reduce labor costs, and reduce errors that may be caused by manual operation. By optimizing the frequency of use and load, the wear and failure rate of the elevator are reduced, and the service life of the equipment is extended.
[0011] Preferably, step S2 specifically includes:
[0012] Perform feature extraction based on the elevator usage data to obtain elevator usage feature data;
[0013] Perform feature classification according to the elevator usage feature data to obtain usage feature classification data;
[0014] Perform feature clustering processing based on feature classification data to obtain feature cluster data;
[0015] Decompose the time segment according to the characteristic cluster data to obtain the characteristic time segment data;
[0016] Generate usage graph nodes and generate usage graph edges for usage feature time segmentation data, and obtain usage graph node data and usage graph edge data respectively;
[0017] The graph is constructed according to the usage graph node data and the usage graph edge data to obtain the construction elevator usage graph data.
[0018] In the present invention, through detailed feature extraction, classification and clustering, the use of the elevator can be understood more accurately, so as to carry out targeted optimization. Data dimension reduction and clustering processing reduce data redundancy, improve the efficiency of calculation and the speed of subsequent analysis. Feature classification and clustering can identify different usage patterns and rules, and provide a basis for optimizing the scheduling and operation strategy of the elevator. Time segment decomposition and graph construction provide support for dynamic optimization, which can be flexibly adjusted according to real-time data, and improve the response speed and adaptability of the system. Graph construction converts complex usage data into an intuitive graph structure, which is convenient for comprehensive analysis and optimization, and helps identify critical paths and optimization bottlenecks.
[0019] Preferably, the feature extraction is specifically:
[0020] Carry out load fluctuation processing according to the elevator usage data to obtain load fluctuation characteristic data;
[0021] Process the usage time period according to the elevator usage data to obtain usage time period data;
[0022] Calculate the number of moving floors according to the elevator usage data to obtain the number of moving floors data;
[0023] Extract the resident layer according to the elevator usage data to obtain the resident layer data;
[0024] The usage frequency is calculated based on the elevator usage data and the load fluctuation characteristic data to obtain the usage frequency data;
[0025] The usage frequency data, load fluctuation characteristic data, usage time period data, moving floor number data and resident floor data are vectorized to obtain the elevator usage characteristic data.
[0026] The extraction of various features in the present invention covers different aspects of elevator use and provides a comprehensive usage analysis. By analyzing load fluctuations, usage time periods, number of moving floors and resident floors, the scheduling strategy and operation path of the elevator can be optimized to improve efficiency. Extracting resident floor data and usage frequency data helps to improve the elevator's response speed to call requests and overall usage efficiency. By optimizing load management and reducing unnecessary movements, the energy consumption and equipment wear of the elevator are reduced, and the service life of the equipment is extended. The analysis of usage frequency and load fluctuations can help predict possible failures, perform preventive maintenance, and reduce unexpected downtime and maintenance costs.
[0027] Preferably, the load fluctuation processing is specifically as follows:
[0028] Extract load data based on elevator usage data to obtain elevator load data;
[0029] Perform spectrum conversion according to the elevator load data to obtain load spectrum data;
[0030] Perform time series segmentation according to the load spectrum data to obtain load spectrum segmentation data;
[0031] The fluctuation amplitude is calculated according to the load spectrum segmentation data to obtain the load fluctuation amplitude data;
[0032] Classify usage scenarios according to load fluctuation amplitude data to obtain classified load data;
[0033] Load fluctuation characteristics are extracted according to the classified load data to obtain load fluctuation characteristic data.
[0034] In the present invention, through spectrum conversion and time segmentation, the changing patterns and periodic characteristics of load data can be deeply analyzed to provide more comprehensive fluctuation information. By calculating the fluctuation amplitude and classifying the usage scenarios, abnormal fluctuations can be identified and classified, which helps to prevent equipment failures and optimize management strategies. Time segmentation and dynamic fluctuation analysis support real-time monitoring and dynamic adaptation, allowing the system to flexibly respond to different usage conditions and load changes. Load fluctuation feature extraction simplifies complex data into key features, provides data-driven decision support, and improves the scientificity and accuracy of decision-making. By optimizing load management and real-time monitoring, the operating efficiency and safety of the elevator system can be improved, the equipment life can be extended, and maintenance costs can be reduced.
[0035] Preferably, the usage frequency calculation is specifically:
[0036] Divide the load interval according to the load fluctuation characteristic data and the elevator usage data to obtain the load interval division data;
[0037] Calculate the load usage based on the load interval division data to obtain the load interval usage frequency data of the elevator;
[0038] Perform usage frequency calculation according to the usage frequency data of the elevator load interval to obtain first usage frequency data;
[0039] Perform event division according to the elevator usage data and preset elevator event parameter data to obtain event division data;
[0040] Calculate event usage based on event classification data to obtain elevator event usage frequency data;
[0041] The usage frequency is calculated according to the elevator event usage frequency data to obtain second usage frequency data.
[0042] In the present invention, frequency calculation is performed through two dimensions, load interval and event division, which can more comprehensively analyze the use of the elevator and provide multi-dimensional optimization suggestions. The detailed classification of load interval division and event division helps to more accurately identify different usage modes and load conditions and optimize the scheduling strategy. The analysis results of load frequency and event frequency can help improve the efficiency of the use of the elevator, reduce energy consumption, and improve safety. Through dynamic calculation and analysis of the usage frequency, real-time monitoring and dynamic adjustment can be achieved to ensure that the elevator is always in the best operating state. The frequency calculation results can help optimize the scheduling and resource allocation of the elevator and improve the overall efficiency and response speed of the construction site.
[0043] Preferably, the load usage calculation is specifically as follows:
[0044] Clustering is performed according to the load interval division data to obtain load interval clustering data;
[0045] Performing Gaussian distribution processing on the load interval clustering data to obtain load interval clustering Gaussian distribution data;
[0046] Perform feature extraction based on the load interval clustered Gaussian distribution data to obtain load interval clustered Gaussian distribution feature data;
[0047] Similarity calculation is performed based on the Gaussian distribution characteristic data of the load interval clusters to obtain cluster similarity data;
[0048] Aggregate the load interval partition data according to the cluster similarity data to obtain load interval aggregate data;
[0049] The usage frequency is calculated based on the load interval aggregation data to obtain the elevator load interval usage frequency data.
[0050] The clustering processing in the present invention can identify the usage patterns of load intervals, simplify the data structure, reduce the computational complexity, and improve the processing efficiency. Gaussian distribution processing smoothes the noise in the data and more accurately describes the data characteristics through probabilistic modeling, providing a solid foundation for subsequent analysis. Feature extraction identifies key features in the data and reduces the amount of calculation through data dimensionality reduction, thereby improving processing speed and efficiency. Similarity calculation measures the similarity of different load intervals, optimizes the clustering results, and ensures the accuracy and rationality of data classification. Load interval aggregation integrates similar data, reduces data redundancy, and improves data consistency and processing efficiency. The frequency of use calculation based on aggregated data is more accurate, which can provide important decision support for elevator scheduling and management, optimize scheduling strategies, and improve usage efficiency.
[0051] Preferably, step S3 specifically includes:
[0052] Calculate the node residence probability and node residence time according to the construction elevator usage graph data to obtain the node residence probability data and node residence time data;
[0053] Extract residence layer features based on node residence probability data and node residence time data to obtain residence feature data;
[0054] Performing clustering processing according to the resident feature data to obtain resident clustering data;
[0055] Perform dynamic calculation of the resident layer according to the resident clustering data to obtain dynamic resident data;
[0056] Construct a resident graph according to the dynamic resident data to obtain the resident graph data of the construction elevator;
[0057] The current construction period data is obtained, and the current construction period data and the construction elevator residence map data are matched to obtain the current construction elevator residence data.
[0058] The present invention identifies high-frequency residence layers and residence durations through residence probability and time calculation, optimizes residence layer settings, and improves elevator scheduling efficiency. Dynamically calculate and adjust residence layer settings so that the system can be optimized in real time according to actual conditions, thereby improving flexibility and adaptability. Through feature extraction and clustering processing, detailed data support is provided to build a more accurate optimization model and improve the scientific nature of decision-making. Optimize residence layer settings and scheduling strategies to reduce elevator waiting time and unnecessary movement, reduce energy consumption, and improve resource utilization efficiency. Dynamically match construction period data to ensure that elevator scheduling is synchronized with construction progress, thereby improving construction efficiency and safety. Through the construction of a residence map, the elevator usage and residence layer settings are intuitively displayed, which facilitates management and monitoring and improves management efficiency.
[0059] Preferably, the resident layer dynamic calculation is specifically:
[0060] Performing feature expansion according to the resident cluster data to obtain resident cluster expansion data;
[0061] Performing short-term trend processing and long-term trend processing according to the resident cluster extension data to obtain resident cluster short-term trend data and resident cluster long-term trend data;
[0062] Generate dynamic weights according to the resident cluster short-term trend data and the resident cluster long-term trend data to obtain resident dynamic weight data;
[0063] According to the resident clustering data and the resident dynamic weight data, dynamic resident data is obtained.
[0064] In the present invention, feature expansion is used to increase the data dimension, making the analysis of the resident layer more comprehensive and detailed, and able to capture more subtle changes. Short-term trend processing can quickly capture the latest change trends, allowing the system to respond in a timely manner, optimize real-time scheduling, and improve system flexibility. Long-term trend processing can identify the long-term changes and stability of the resident layer settings, provide a basis for long-term optimization and preventive maintenance of the system, and extend the service life of the equipment. Dynamic weight generation enables the system to flexibly adjust the resident layer settings according to the needs of different time periods, optimize resource allocation, reduce waiting time and energy consumption, and improve overall efficiency. Dynamic resident data comprehensively considers short-term and long-term trends, making the resident layer settings more accurate and comprehensive, supporting real-time adjustments, and improving system response speed and flexibility.
[0065] Preferably, step S4 is specifically:
[0066] Obtain current elevator usage data, and extract current resident layer data according to the current elevator usage data to obtain current resident layer data;
[0067] Process the change trend according to the current resident layer data to obtain the resident layer trend data;
[0068] The residence optimization is performed according to the residence layer trend data and the construction elevator residence recommendation data, and the current construction elevator residence optimization data is obtained to perform intelligent control operations of the construction elevator.
[0069] In the present invention, by real-time acquisition and processing of current elevator usage data, the system can dynamically adapt to the actual situation of the construction site, make timely adjustments, and improve the flexibility and response speed of the system. Change trend processing can identify the change trend of the current resident layer data, foresee possible problems in advance, make preventive adjustments, and improve the stability and reliability of the system. Combining trend data and resident recommendation data for resident optimization can comprehensively consider short-term and long-term factors, provide a more comprehensive optimization plan, and improve the scientificity and accuracy of decision-making. By optimizing the resident layer setting, the waiting time and unnecessary movement of the elevator can be reduced, work efficiency can be improved, energy consumption and equipment wear can be reduced, and equipment life can be extended. Resident optimization helps to improve the scheduling efficiency of the elevator, ensure the smooth progress of the construction progress, while reducing equipment failures and improving overall construction safety.
[0070] Preferably, the present application also provides a construction elevator intelligent control system, which is used to execute the construction elevator intelligent control method as described above, and the construction elevator intelligent control system includes:
[0071] The elevator basic data collection module is used to obtain elevator usage data and construction period data;
[0072] A construction elevator usage graph construction module is used to extract usage features based on elevator usage data to obtain elevator usage feature data, and to construct a construction elevator usage graph based on the elevator usage feature data and construction period data to obtain construction elevator usage graph data;
[0073] The construction elevator residence recommendation module is used to calculate the residence layer according to the construction elevator usage map data to obtain the construction elevator residence map data; obtain the current construction period data, and match the current construction period data and the construction elevator residence map data to obtain the construction elevator residence recommendation data;
[0074] The current construction elevator residence optimization module is used to obtain the current elevator usage data, perform residence optimization based on the current elevator usage data and the construction elevator residence recommendation data, and obtain the current construction elevator residence optimization data to perform intelligent control operations on the construction elevator.
[0075] The beneficial effects of the present invention are as follows: by combining global feature extraction with local resident optimization, global optimization and local adaptation of the system are achieved, ensuring efficient operation at different construction stages. Through real-time data acquisition and dynamic weight generation, the system can make real-time adjustments according to actual conditions to ensure the flexibility and response speed of the system. Comprehensive multi-dimensional data analysis, combined with short-term and long-term trends, provides scientific decision-making support to ensure efficient operation of the system in complex construction environments. By optimizing scheduling and resident layer settings, the waiting time and unnecessary operation of the elevator are reduced, energy consumption is reduced, energy conservation and emission reduction are achieved, and resource utilization efficiency is improved. Through change trend analysis and preventive adjustments, potential problems can be identified in advance, preventive maintenance can be carried out, equipment failures can be reduced, and the safety and reliability of the system can be ensured. BRIEF DESCRIPTION OF THE DRAWINGS
[0076] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting implementations made with reference to the following drawings:
[0077] Figure 1 A flowchart showing a method for intelligently controlling a construction elevator according to an embodiment of the present invention is provided;
[0078] Figure 2 A flowchart showing the steps of a method for constructing a construction hoist usage diagram according to an embodiment is shown;
[0079] Figure 3 A flowchart showing a construction elevator residence recommendation method according to an embodiment of the present invention;
[0080] Figure 4 A flowchart of the steps of a current construction elevator residence optimization method according to an embodiment is shown. DETAILED DESCRIPTION
[0081] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.
[0082] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.
[0083] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.
[0084] See also Figures 1 to 4 , the present application provides a construction hoist intelligent control method, comprising the following steps:
[0085] Step S1: Obtaining elevator usage data and construction period data;
[0086] Specifically, sensors and IoT devices are used to collect the usage data of the elevator in real time, including but not limited to the start time, stop time, operation time, load, etc. The collected usage data is uploaded to the cloud database for persistent storage. The construction period data is obtained from the construction project management system, including the project start date, expected completion date, completion status of tasks at each stage, etc.
[0087] Step S2: extracting usage features based on the elevator usage data to obtain elevator usage feature data, and constructing a construction elevator usage graph based on the elevator usage feature data and the construction period data to obtain construction elevator usage graph data;
[0088] Specifically, feature extraction is performed based on the elevator usage data. The specific method includes: calculating the average daily usage time and peak usage period of the elevator. Calculating the average load, maximum load, load change trend, etc. of the elevator. Counting the start frequency, stop frequency, continuous usage time, etc. of the elevator. The extracted feature data is standardized to generate elevator usage feature data.
[0089] Based on the elevator usage feature data and construction period data, a construction elevator usage graph is constructed. The specific method includes: each node represents the elevator usage in a specific time period. The edges between nodes represent the usage association between different time periods, and the weight of the edge is determined by calculating the similarity or correlation between time periods. The information of nodes and edges is integrated to generate a complete construction elevator usage graph.
[0090] Step S3: Calculate the resident layer according to the construction elevator usage map data to obtain the construction elevator resident map data; obtain the current construction period data, and match the current construction period data and the construction elevator resident map data to obtain the construction elevator resident recommendation data;
[0091] Specifically, based on the construction elevator usage graph data, the residence of the elevator on each floor is calculated. The specific method includes: counting the elevator's stay time on each floor, calculating the average residence time and residence frequency. Using clustering algorithms (such as K-means or DBSCAN), the residence pattern of the elevator on each floor is identified to find the permanent and short-term residence layers.
[0092] Obtain data from the current construction phase from the construction project management system to determine the current construction tasks and requirements. Match the current construction period data with the construction elevator residence map data. The specific method includes: analyzing the demand for elevator residence in the current construction phase, determining the priority residence layer and the backup residence layer. Calculate the optimal residence layer allocation plan through optimization algorithms (such as linear programming or genetic algorithms) to generate construction elevator residence recommendation data.
[0093] Step S4: Acquire current elevator usage data, perform residence optimization based on the current elevator usage data and construction elevator residence recommendation data, and obtain current construction elevator residence optimization data to perform intelligent control operations on the construction elevator.
[0094] Specifically, the current elevator usage data is collected in real time, including the current operating floor, operating time, load, etc.
[0095] Based on the current elevator usage data and the recommended data for construction elevator residence, residence optimization is performed. The specific method includes: matching the current usage data with the recommended data to determine whether the residence layer needs to be adjusted. Using heuristic algorithms or deep learning models, the optimal residence layer adjustment plan is calculated to ensure the efficiency and safety of elevator use.
[0096] According to the residence optimization data, intelligent control operations are carried out. Specific methods include: controlling the elevator to automatically adjust the residence layer through the automatic scheduling system to ensure that the construction needs are met. Real-time monitoring of the operation status of the elevator, detection of abnormal conditions, and timely warning and processing.
[0097] Specifically, in a 90-day construction project, the usage data of the elevator are as follows: start time: 7:00-17:00 every day, stop time: 17:00-7:00 every day, average daily operating time: 8 hours, load: average load 500kg, maximum load 1000kg.
[0098] Construction period data: Project start date: July 1, 2023, expected completion date: September 28, 2023, current construction stage: foundation stage, requiring a large amount of material transportation.
[0099] The extracted characteristic data are as follows: average daily usage time: 8 hours, peak usage period: 10:00-12:00, 14:00-16:00, average load: 500kg, maximum load: 1000kg, start frequency: 10 times / day, stop frequency: 10 times / day.
[0100] Use graph construction (each node represents the usage of the elevator in a specific time period. The edges between nodes represent the usage association between different time periods, and the weight of the edge is determined by calculating the similarity or correlation between time periods.): Node: represents the usage data of each day, such as the 1st day, the 2nd day, etc. Edge: represents the usage association between adjacent days, such as the similarity between the 1st day and the 2nd day.
[0101] The residence layer is calculated as follows (the elevator's stay time on each layer is counted, and the average residence time and residence frequency are calculated. The residence pattern of the elevator on each layer is identified through a clustering algorithm (such as K-means or DBSCAN), and the permanent layer and the short-term residence layer are found.): The residence time on the first layer: 4 hours, the residence frequency: 5 times, the residence time on the second layer: 3 hours, the residence frequency: 4 times, the residence time on the third layer: 1 hour, the residence frequency: 1 time.
[0102] Residency pattern recognition results: permanent residence layer: layer 1, layer 2, short-term residence layer: layer 3. Current construction phase requirements: materials need to be frequently transported to layers 1 and 2. Residency recommendation data: priority residence layer: layer 1, backup residence layer: layer 2.
[0103] Current operating floor: Floor 1, current load: 600kg. Optimized stay calculation: Keep the elevator on Floor 1 based on the current load and the recommendation of the preferred stay floor. Intelligent control operation: Automatic scheduling system instructions: Keep the elevator on Floor 1 to ensure efficient transportation of construction materials.
[0104] Preferably, step S2 specifically includes:
[0105] Step S21: extracting features based on the elevator usage data to obtain elevator usage feature data;
[0106] Specifically, relevant features are extracted from the elevator usage data, such as running time, load, frequency, etc. Data is collected from the elevator monitoring system, including: timestamp, load, running time, stop floors, up and down, etc.
[0107] Extract the running time characteristics of the elevator, that is, the time difference between the start and stop of each elevator operation. Extract the timestamps of the elevator start and stop from the data. Use the time difference to calculate the duration of each elevator operation: perform statistical calculations on the running time to obtain running time characteristic data (such as mean, maximum, minimum, standard deviation, etc.);
[0108] Extract the load characteristics of the elevator and analyze the average load, maximum load and fluctuation of the elevator during each operation. Extract the load information of each elevator operation from the data. Analyze the load data of each operation and calculate the average load, maximum load, minimum load and load fluctuation (standard deviation): Calculate the overall distribution characteristics of the load and obtain the operation time characteristic data (such as the mean, maximum, minimum and fluctuation of the load);
[0109] Extract the operating frequency characteristics of the elevator and analyze the number of times the elevator is started within a given time range (such as every day or every hour). Extract the start timestamp of each elevator operation. Calculate the number of elevator operations by time period (such as hours, days, weeks); calculate the frequency statistics to obtain the operating frequency characteristic data (such as average operating frequency, maximum operating frequency, and minimum operating frequency);
[0110] The operating time characteristic data, the operating time characteristic data and the operating frequency characteristic data are integrated to obtain the elevator usage characteristic data.
[0111] Step S22: performing feature classification according to the elevator usage feature data to obtain usage feature classification data;
[0112] Specifically, the extracted feature data are classified, such as time features, load features, number of layers features, etc.
[0113] Classify the time-related features in the extracted feature data. Identify time-related features, such as the use time, use frequency, peak time period, etc. of the elevator. Extract time feature data from the feature data set to form a time feature subset. Analyze the time feature data according to different time periods, such as morning peak, evening peak, night, etc. Use a classification algorithm (such as K-means clustering or hierarchical clustering) to classify the time features and obtain the usage feature classification data for different time periods.
[0114] K-means clustering algorithm is used for time feature classification: 1. Randomly select K initial cluster centers. 2. Calculate the distance from each time feature data point to each cluster center, and assign each data point to the cluster center with the closest distance. 3. Calculate the new center of each cluster, that is, the average value of all data points assigned to the cluster. 4. Repeat steps 2 and 3 until the cluster center no longer changes or the maximum number of iterations is reached. Output result: K clusters are obtained, each cluster corresponds to a time feature classification data.
[0115] Classify the features related to load in the extracted feature data. Identify features related to load, such as load fluctuation, average load, maximum load, minimum load, etc. Extract load feature data from the feature data set to form a load feature subset. Divide the load feature data into different load ranges, such as 0-500kg, 500-1000kg, etc. Use a classification algorithm (such as a Gaussian mixture model or a decision tree) to classify the load features and obtain the usage feature classification data for different load ranges.
[0116] The Gaussian mixture model is used for load feature classification as follows: 1. Set the initial parameters of the Gaussian mixture model (mean, covariance matrix and mixing coefficient). 2. Calculate the probability (responsibility value) that each load feature data point belongs to each Gaussian distribution. 3. Update the parameters of the Gaussian mixture model to maximize the log-likelihood function. 4. Repeat steps 2 and 3 until the parameters converge or the maximum number of iterations is reached. Output result: multiple Gaussian distributions are obtained, each corresponding to a load feature classification data.
[0117] Classify the features related to the number of floors in the extracted feature data. Identify features related to the number of floors, such as the number of mobile floors, the number of resident floors, the frequency of floor changes, etc. Extract the floor feature data from the feature data set to form a floor feature subset. Divide the floor feature data by different floors, such as low-rise, middle-rise, high-rise, etc. Use a classification algorithm (such as random forest or support vector machine) to classify the floor features and obtain the usage feature classification data of different floors.
[0118] Random forest is used for layer feature classification, such as building multiple decision trees, each tree uses a random subset of the training data. For each subset, the decision tree is trained using the layer feature data, and the tree structure is constructed by selecting the optimal split point. The new layer feature data is input into each decision tree for classification. Using a voting mechanism, the classification results of all decision trees are voted to determine the final classification result. The classification data of the layer feature is obtained.
[0119] Step S23: performing feature clustering processing based on the feature classification data to obtain feature cluster data;
[0120] Specifically, a clustering algorithm (such as K-means) is used to cluster the feature classification data to obtain feature cluster data. 1. Randomly select K (such as 3, 5, 7 or a preset value) initial cluster centers. 2. Calculate the distance from each feature classification data to each cluster center, and assign each data point to the cluster center with the nearest distance. 3. Calculate the new center of each cluster, that is, the average of all data points assigned to the cluster. 4. Repeat steps 2 and 3 until the cluster center no longer changes or the maximum number of iterations is reached. Output result: K clusters are obtained, each cluster corresponds to a feature cluster data.
[0121] Step S24: performing time segment decomposition according to the feature cluster data to obtain the use feature time segment data;
[0122] Specifically, segment the feature cluster data by time (fixed time period segmentation). Determine the segmentation rules for the time period, such as by hour, day, week or other time intervals. Alternatively, set a fixed time interval (working time or importing duration data) to ensure that the feature data can be segmented according to the same time span. According to the time segmentation rules, group the feature cluster data by time period. Assign a corresponding time period to each data point. For example, for data segmented every day, map each data point to the corresponding date. Group the data by time period, and each segment contains all the feature cluster data within the time period.
[0123] Step S25: generating usage graph nodes and usage graph edges for the usage feature time segment data, and obtaining usage graph node data and usage graph edge data respectively;
[0124] Specifically, the data of each time segment is taken as a node, and the edges of the graph are generated according to the data relationship of adjacent time segments.
[0125] Step S26: construct a graph based on the usage graph node data and the usage graph edge data to obtain the construction elevator usage graph data.
[0126] Specifically, based on the generated nodes and edges, a construction elevator usage graph is constructed.
[0127] Preferably, the feature extraction is specifically:
[0128] Carry out load fluctuation processing according to the elevator usage data to obtain load fluctuation characteristic data;
[0129] Specifically, the load data of each operation of the lift is obtained from the sensor. The load variation during each operation is calculated, and the load fluctuation characteristics are described using the standard deviation or variance.
[0130] Process the usage time period according to the elevator usage data to obtain usage time period data;
[0131] Specifically, record the daily operation time period of the elevator. Count the start and stop time of the elevator every day, and calculate the usage of each time period. The collected usage time data: 7:00-9:00, 10:00-12:00, 14:00-16:00, 17:00-19:00. Calculate the usage time period data: usage time period = {7:00-9:00, 10:00-12:00, 14:00-16:00, 17:00-19:00}.
[0132] Calculate the number of moving floors according to the elevator usage data to obtain the number of moving floors data;
[0133] Specifically, record the number of floors the elevator moves in a day. Count the number of floors the elevator moves every day. Collected floor data: 1st floor - 5th floor, 5th floor - 10th floor, 10th floor - 15th floor, 15th floor - 20th floor; Calculate the number of floors moved: number of floors moved = 20th floor - 1st floor = 19th floor.
[0134] Extract the resident layer according to the elevator usage data to obtain the resident layer data;
[0135] Specifically, record the time the elevator stays on each floor. Count the time the elevator stays on each floor and extract the permanent floor. The collected residence time data: 1st floor (2 hours), 5th floor (1 hour), 10th floor (3 hours), 15th floor (4 hours), 20th floor (1 hour). Calculate the residence floor data: permanent floor = 15th floor (4 hours). According to the statistical residence time, find the floor with the longest residence time as the permanent floor.
[0136] The usage frequency is calculated based on the elevator usage data and the load fluctuation characteristic data to obtain the usage frequency data;
[0137] Specifically, the use frequency of the elevator is counted by combining the use time period data and the load fluctuation characteristic data. The use frequency can be calculated by the number of starts per day.
[0138] The usage frequency data, load fluctuation characteristic data, usage time period data, moving floor number data and resident floor data are vectorized to obtain the elevator usage characteristic data.
[0139] Specifically, the usage frequency data, load fluctuation characteristic data, usage time period data, moving layer number data and resident layer data are integrated into a feature vector.
[0140] Preferably, the load fluctuation processing is specifically as follows:
[0141] Extract load data based on elevator usage data to obtain elevator load data;
[0142] Specifically, the load data of the elevator, including the load weight during each operation, is recorded in real time through sensors installed on the elevator.
[0143] Perform spectrum conversion according to the elevator load data to obtain load spectrum data;
[0144] Specifically, a fast Fourier transform (FFT) is performed on the load data to convert the time domain data into frequency domain data to obtain the load spectrum data.
[0145] Perform time series segmentation according to the load spectrum data to obtain load spectrum segmentation data;
[0146] Specifically, the load spectrum data is segmented by time for further analysis. {"Period 1": [(5270+0j), (20-72.62j)],"Period 2": [(30+12.36j), (40-40j)],"Period 3": [(30-20j), (-30+0j)],"Period 4": [(30+20j), (40+40j)],"Period 5": [(30-12.36j), (20+72.62j)]}.
[0147] The fluctuation amplitude is calculated according to the load spectrum segmentation data to obtain the load fluctuation amplitude data;
[0148] Specifically, calculate the fluctuation amplitude of the load spectrum data in each period, using the amplitude formula: , is the load fluctuation amplitude data, is the real part of the load spectrum segment data, It is the imaginary part of the load spectrum segment data.
[0149] Classify usage scenarios according to load fluctuation amplitude data to obtain classified load data;
[0150] Specifically, based on the fluctuation data, clustering algorithms (such as K-means) are used to classify the data and determine different usage scenarios. After multiple iterations, K stable clusters are obtained, each of which corresponds to a usage scenario. Classified load data is formed, and each data point is assigned to the corresponding cluster to identify its usage scenario.
[0151] Assign a label to each data point to indicate the cluster it belongs to. For example, if there are K clusters, the label can be an integer from 1 to K. Record each data point and its label in the classification load data set to form the classification result.
[0152] Specifically, the time window load data is classified according to different usage scenarios of the construction site, such as peak hours, nighttime construction, etc. Combined with the construction plan and schedule, the load data is assigned to different usage scenario categories.
[0153] Load fluctuation characteristics are extracted according to the classified load data to obtain load fluctuation characteristic data.
[0154] Specifically, features are extracted from the classified load data, including average amplitude, maximum amplitude, fluctuation frequency, etc.
[0155] Preferably, the usage frequency calculation is specifically:
[0156] Divide the load interval according to the load fluctuation characteristic data and the elevator usage data to obtain the load interval division data;
[0157] Specifically, the load data is divided into different intervals according to the load fluctuation characteristic data and the elevator usage data. The load intervals are divided (eg, one interval for every 50 kg) to obtain load interval division data.
[0158] Calculate the load usage based on the load interval division data to obtain the load interval usage frequency data of the elevator;
[0159] Specifically, the usage frequency in each load interval is counted.
[0160] Perform usage frequency calculation according to the usage frequency data of the elevator load interval to obtain first usage frequency data;
[0161] Specifically, the first usage frequency of the elevator is calculated based on the load interval usage frequency data using a frequency calculation formula.
[0162] Perform event division according to the elevator usage data and preset elevator event parameter data to obtain event division data;
[0163] Specifically, events are divided according to elevator usage data and preset elevator event parameter data (such as start, stop, failure, etc.).
[0164] Calculate event usage based on event classification data to obtain elevator event usage frequency data;
[0165] Specifically, count the frequency of use of each event. Collect event data for the elevator. The event data should include the type of each event (such as "start", "stop", "overload", etc.), timestamp, and other related information. Classify each event to ensure that each event corresponds to a category (such as "start", "stop", "overload", etc.). Extract event type information from the event data. Divide events into different categories. Count each event type and calculate the frequency of each event. Use Pandas' groupby and count methods to count the number of times each type of event occurs.
[0166] The usage frequency is calculated according to the elevator event usage frequency data to obtain second usage frequency data.
[0167] Specifically, according to the event usage frequency data, the frequency calculation formula is used to calculate the second usage frequency of the elevator. The frequency calculation formula is specifically: , is the second usage frequency data, Frequency data for events. The duration of the time period.
[0168] Preferably, the load usage calculation is specifically as follows:
[0169] Clustering is performed according to the load interval division data to obtain load interval clustering data;
[0170] Specifically, clustering algorithms such as K-means are used to cluster the load interval data to obtain load interval clustering data. The K-means clustering algorithm is used to divide the load interval data: 1. Randomly select K initial cluster centers. 2. Calculate the distance from each time feature data point to each cluster center, and assign each data point to the cluster center with the nearest distance. 3. Calculate the new center of each cluster, that is, the average of all data points assigned to the cluster. 4. Repeat steps 2 and 3 until the cluster center no longer changes or the maximum number of iterations is reached. Combine the cluster labels with the original data to generate a complete load interval clustering data set, including the original load interval data and the corresponding cluster labels, to obtain the load interval clustering data.
[0171] Performing Gaussian distribution processing on the load interval clustering data to obtain load interval clustering Gaussian distribution data;
[0172] Specifically, Gaussian distribution fitting is performed on the clustered data to obtain Gaussian distribution parameters (mean and variance). Core features (such as mean and variance) are extracted from the Gaussian distribution parameters as the distribution characteristics of the clustered load. The Gaussian distribution mean of each cluster is extracted as a feature. The mean reflects the central load value of the cluster. The Gaussian distribution variance (square of the standard deviation) of each cluster is extracted as a feature. The variance represents the volatility of the load. The larger the variance, the more dispersed the load distribution. The kurtosis of the load distribution in the cluster is extracted to indicate whether the distribution is peaked or flat. The skewness of the load distribution is extracted to indicate whether the distribution is symmetrical or tilted. The confidence intervals of the mean and variance are calculated to further characterize the distribution range of the load. All extracted features (such as mean, variance, kurtosis, skewness, etc.) are integrated into a data set to form complete load interval clustering Gaussian distribution feature data.
[0173] Perform feature extraction based on the load interval clustered Gaussian distribution data to obtain load interval clustered Gaussian distribution feature data;
[0174] Specifically, features such as mean, variance, etc. are extracted from Gaussian distribution parameters. The Gaussian distribution mean of each cluster is extracted, and the mean represents the central value of the load in the cluster. These mean features are analyzed to extract the load distribution center of each cluster and determine which clusters have load values concentrated in which intervals.
[0175] The variance is extracted from the Gaussian distribution parameters. This feature describes the volatility of the load distribution. The variance is the square of the standard deviation, so it can be calculated by squaring the standard deviation. The larger the variance, the greater the load volatility; the smaller the variance, the smaller the load volatility. By analyzing the variance, clusters with stable loads and clusters with large load fluctuations are identified. All extracted features are integrated into a data set to generate complete load interval clustering Gaussian distribution feature data.
[0176] Similarity calculation is performed based on the load interval clustering Gaussian distribution characteristic data to obtain clustering similarity data;
[0177] Specifically, the similarity between each cluster feature is calculated using methods such as Euclidean distance. Before calculating the similarity, the features are standardized to ensure that all features have the same scale to avoid unnecessary influence of a feature on the similarity calculation due to its large value. The similarity between each cluster feature is calculated using Euclidean distance, which is a method for measuring the distance between two points in a multidimensional space. The cdist function in the scipy library is used to calculate the Euclidean distance for the feature vectors of each two clusters to obtain a Euclidean distance matrix. Each element in the Euclidean distance matrix represents the Euclidean distance between two clusters. The smaller the value, the more similar the two clusters are, and the larger the value, the greater the difference between the two clusters. The Euclidean distance matrix is converted into structured data to generate a cluster similarity dataset. Similarity can be represented by calculating the opposite number (for example, 1 / distance). The larger the value, the more similar the clusters are.
[0178] Aggregate the load interval partition data according to the cluster similarity data to obtain load interval aggregate data;
[0179] Specifically, similar clusters are aggregated based on similarity data. A similarity threshold is set to determine which clusters can be aggregated. Clusters with high similarity can be aggregated into a new cluster. A similarity threshold is set based on actual needs. Clusters with similarity above this threshold will be aggregated. A similarity value close to 1 indicates that the clusters are very similar, and a similarity value close to 0 indicates that the difference is large. Set a similarity threshold, such as 0.8 or above, to aggregate clusters. Screen the similarity matrix to find all cluster pairs whose similarity is greater than the set threshold.
[0180] Based on the similarity data, similar clusters are merged to generate new cluster labels. A mapping table is created for similar cluster pairs, and similar clusters are mapped to a new cluster label. The original cluster label is updated to the new cluster label to generate new load interval aggregate data. The aggregated load interval data will have the updated cluster label, and the load interval aggregate data will be obtained.
[0181] The usage frequency is calculated based on the load interval aggregation data to obtain the elevator load interval usage frequency data.
[0182] Specifically, the usage frequency of each interval is calculated based on the aggregated load interval data. Define the rules for calculating the usage frequency of each interval, and count the usage times based on time periods (such as hours, days, and weeks). Assuming that there is timestamp information in the data, first divide it by time period, such as by day, by hour, etc. If there is no timestamp, define the usage times in other ways. Alternatively, count by time (such as every day, every hour), or count the usage frequency of each interval by running interval.
[0183] According to the defined frequency calculation rules, count the number of times each load interval is used within the set time period. For each time period, count the usage frequency of each load interval (or cluster). If it is not subdivided by time period, you can also directly calculate the total usage frequency of each interval. Integrate the usage frequency data into a new data set, integrate the usage frequency of each cluster into a load interval usage frequency data set, and generate load interval usage frequency data.
[0184] Preferably, step S3 specifically includes:
[0185] Step S31: Calculate the node residence probability and the node residence time according to the construction elevator usage graph data to obtain the node residence probability data and the node residence time data;
[0186] Specifically, based on the construction elevator usage graph data, the residence probability of each node is calculated, and the residence time of each node is calculated.
[0187] Step S32: extracting residence layer features according to the node residence probability data and the node residence time data to obtain residence feature data;
[0188] Specifically, residence features are extracted from node residence probability data and residence time data.
[0189] Step S33: performing clustering processing according to the resident feature data to obtain resident cluster data;
[0190] Specifically, clustering (such as K-means algorithm) is performed on the resident feature data to obtain resident cluster data.
[0191] Step S34: Perform dynamic calculation of the resident layer according to the resident cluster data to obtain dynamic resident data;
[0192] Specifically, the dynamic resident data of each resident layer is calculated based on the resident cluster data, such as calculating the resident layer data at different levels or under different classification conditions based on the classification effect of cluster calculation.
[0193] Step S35: constructing a resident graph according to the dynamic resident data to obtain resident graph data of the construction elevator;
[0194] Specifically, node data representing the floor is generated based on the dynamic resident data. Extract floor information: extract the resident time and resident frequency of each floor from the dynamic resident data. Define each floor as a node, and the node attributes include floor number, resident time, and resident frequency. Assign corresponding dynamic resident data attributes to each node to form a node data set.
[0195] Generate edge data representing inter-floor movement based on dynamic resident data. Extract the number of inter-floor movement and movement time from the dynamic resident data. Define the movement between each two adjacent floors as an edge, and the attributes of the edge include the number of movement and movement time. Assign corresponding dynamic resident data attributes to each edge to form an edge data set.
[0196] Based on the node data and edge data, construct a construction elevator resident graph. Use a graph structure library (such as NetworkX or Graphviz) to initialize an empty graph. Add each node in the node dataset to the graph and set the node attributes. Add each edge in the edge dataset to the graph and set the edge attributes. Adjust the layout of the graph to make the graph structure clearer and more intuitive for easy analysis and presentation.
[0197] Step S36: Acquire the current construction period data, and match the current construction period data and the construction elevator residence map data to obtain the current construction elevator residence data.
[0198] Specifically, the dynamic resident data of each resident layer is calculated according to the resident clustering data.
[0199] Preferably, the resident layer dynamic calculation is specifically:
[0200] Performing feature expansion according to the resident cluster data to obtain resident cluster expansion data;
[0201] Specifically, according to the resident clustering data, additional features such as time, frequency, etc. are added to expand the resident feature data.
[0202] Performing short-term trend processing and long-term trend processing according to the resident cluster extension data to obtain resident cluster short-term trend data and resident cluster long-term trend data;
[0203] Specifically, short-term trend processing: using methods such as sliding average to calculate the short-term trend of resident features. Long-term trend processing: using methods such as exponential smoothing to calculate the long-term trend of resident features.
[0204] Generate dynamic weights according to the resident cluster short-term trend data and the resident cluster long-term trend data to obtain resident dynamic weight data;
[0205] Specifically, dynamic weights are generated based on short-term trend data and long-term trend data, such as , To store dynamic weight data, is the weight data corresponding to the short-term trend data, For short-term trend data, is the weight data corresponding to the long-term trend data, Long-term trend data.
[0206] According to the resident clustering data and the resident dynamic weight data, dynamic resident data is obtained.
[0207] Specifically, the dynamic resident data is calculated by combining the resident clustering data and the dynamic weight data.
[0208] Preferably, step S4 is specifically:
[0209] Step S41: obtaining current elevator usage data, and extracting current resident layer data according to the current elevator usage data to obtain current resident layer data;
[0210] Specifically, the current elevator usage data is obtained in real time, including the operating floor, residence time, load, etc. The current residence floor data is extracted according to the current elevator usage data.
[0211] Step S42: performing change trend processing according to the current resident layer data to obtain resident layer trend data;
[0212] Specifically, a time series analysis method (such as sliding average, exponential smoothing, etc.) is used to calculate the changing trend of the current resident layer.
[0213] Step S43: Perform residence optimization according to the residence layer trend data and the construction elevator residence recommendation data to obtain the current construction elevator residence optimization data for performing intelligent control operations of the construction elevator.
[0214] Specifically, based on the residence layer trend data and the construction elevator residence recommendation data, an optimization algorithm (such as linear programming, genetic algorithm, etc.) is used to perform residence optimization.
[0215] Preferably, the present application also provides a construction elevator intelligent control system, which is used to execute the construction elevator intelligent control method as described above, and the construction elevator intelligent control system includes:
[0216] The elevator basic data collection module is used to obtain elevator usage data and construction period data;
[0217] A construction elevator usage graph construction module is used to extract usage features based on elevator usage data to obtain elevator usage feature data, and to construct a construction elevator usage graph based on the elevator usage feature data and construction period data to obtain construction elevator usage graph data;
[0218] The construction elevator residence recommendation module is used to calculate the residence layer according to the construction elevator usage map data to obtain the construction elevator residence map data; obtain the current construction period data, and match the current construction period data and the construction elevator residence map data to obtain the construction elevator residence recommendation data;
[0219] The current construction elevator residence optimization module is used to obtain the current elevator usage data, perform residence optimization based on the current elevator usage data and the construction elevator residence recommendation data, and obtain the current construction elevator residence optimization data to perform intelligent control operations on the construction elevator.
[0220] Therefore, from any point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the attached application documents rather than the above description, and it is intended that all changes falling within the meaning and scope of equivalent elements of the application documents are included in the present invention.
[0221] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.
Claims
1. A construction hoist intelligent control method, characterized in that: The following steps are involved: Step S1: Obtaining elevator usage data and construction period data; Step S2: extracting usage features based on the elevator usage data to obtain elevator usage feature data, and constructing a construction elevator usage graph based on the elevator usage feature data and the construction period data to obtain construction elevator usage graph data; Step S3: Calculate the resident layer according to the construction elevator usage map data to obtain the construction elevator resident map data; obtain the current construction period data, and match the current construction period data and the construction elevator resident map data to obtain the construction elevator resident recommendation data; Step S4: Acquire current elevator usage data, perform residence optimization based on the current elevator usage data and construction elevator residence recommendation data, and obtain current construction elevator residence optimization data to perform intelligent control operations on the construction elevator.
2. The intelligent control method for a construction elevator according to claim 1, characterized in that: Step S2 is specifically as follows: Perform feature extraction based on the elevator usage data to obtain elevator usage feature data; Perform feature classification according to the elevator usage feature data to obtain usage feature classification data; Perform feature clustering processing based on feature classification data to obtain feature cluster data; Decompose the time segment according to the characteristic cluster data to obtain the characteristic time segment data; Generate usage graph nodes and generate usage graph edges for usage feature time segmentation data, and obtain usage graph node data and usage graph edge data respectively; The graph is constructed according to the usage graph node data and the usage graph edge data to obtain the construction elevator usage graph data.
3. The intelligent control method for a construction elevator according to claim 2, characterized in that: The feature extraction is as follows: Carry out load fluctuation processing according to the elevator usage data to obtain load fluctuation characteristic data; Process the usage time period according to the elevator usage data to obtain usage time period data; Calculate the number of moving floors according to the elevator usage data to obtain the number of moving floors data; Extract the resident layer according to the elevator usage data to obtain the resident layer data; The usage frequency is calculated based on the elevator usage data and the load fluctuation characteristic data to obtain the usage frequency data; The usage frequency data, load fluctuation characteristic data, usage time period data, moving floor number data and resident floor data are vectorized to obtain the elevator usage characteristic data.
4. The intelligent control method for a construction elevator according to claim 3 is characterized in that: The specific load fluctuation processing is as follows: Extract load data based on elevator usage data to obtain elevator load data; Perform spectrum conversion according to the elevator load data to obtain load spectrum data; Perform time series segmentation according to the load spectrum data to obtain load spectrum segmentation data; The fluctuation amplitude is calculated according to the load spectrum segmentation data to obtain the load fluctuation amplitude data; Classify usage scenarios according to load fluctuation amplitude data to obtain classified load data; Load fluctuation characteristics are extracted according to the classified load data to obtain load fluctuation characteristic data.
5. The intelligent control method for construction hoist according to claim 3 is characterized in that: The frequency calculation is as follows: Divide the load interval according to the load fluctuation characteristic data and the elevator usage data to obtain the load interval division data; Calculate the load usage based on the load interval division data to obtain the load interval usage frequency data of the elevator; Perform usage frequency calculation according to the usage frequency data of the elevator load interval to obtain first usage frequency data; Perform event division according to the elevator usage data and preset elevator event parameter data to obtain event division data; Calculate event usage based on event classification data to obtain elevator event usage frequency data; The usage frequency is calculated according to the elevator event usage frequency data to obtain second usage frequency data.
6. The intelligent control method for a construction elevator according to claim 5, characterized in that: The load usage calculation is as follows: Clustering is performed according to the load interval division data to obtain load interval clustering data; Performing Gaussian distribution processing on the load interval clustering data to obtain load interval clustering Gaussian distribution data; Perform feature extraction based on the load interval clustered Gaussian distribution data to obtain load interval clustered Gaussian distribution feature data; Similarity calculation is performed based on the Gaussian distribution characteristic data of the load interval clusters to obtain cluster similarity data; Aggregate the load interval partition data according to the cluster similarity data to obtain load interval aggregate data; The usage frequency is calculated based on the load interval aggregation data to obtain the elevator load interval usage frequency data.
7. The intelligent control method for a construction elevator according to claim 1, characterized in that: Step S3 is specifically as follows: Calculate the node residence probability and node residence time according to the construction elevator usage graph data to obtain the node residence probability data and node residence time data; Extract residence layer features based on node residence probability data and node residence time data to obtain residence feature data; Performing clustering processing according to the resident feature data to obtain resident clustering data; Perform dynamic calculation of the resident layer according to the resident clustering data to obtain dynamic resident data; Construct a resident graph according to the dynamic resident data to obtain the resident graph data of the construction elevator; The current construction period data is obtained, and the current construction period data and the construction elevator residence map data are matched to obtain the construction elevator residence recommendation data.
8. The intelligent control method for a construction elevator according to claim 7, characterized in that: The dynamic calculation of the resident layer is as follows: Performing feature expansion according to the resident cluster data to obtain resident cluster expansion data; Performing short-term trend processing and long-term trend processing according to the resident cluster extension data to obtain resident cluster short-term trend data and resident cluster long-term trend data; Generate dynamic weights according to the resident cluster short-term trend data and the resident cluster long-term trend data to obtain resident dynamic weight data; According to the resident clustering data and the resident dynamic weight data, dynamic resident data is obtained.
9. The intelligent control method for a construction elevator according to claim 1, characterized in that: Step S4 is specifically as follows: Obtain current elevator usage data, and extract current resident layer data according to the current elevator usage data to obtain current resident layer data; Process the change trend according to the current resident layer data to obtain the resident layer trend data; The residence optimization is performed according to the residence layer trend data and the construction elevator residence recommendation data, and the current construction elevator residence optimization data is obtained to perform intelligent control operations of the construction elevator.
10. An intelligent control system for a construction hoist, characterized in that: Used to execute the construction hoist intelligent control method according to claim 1, the construction hoist intelligent control system comprises: The elevator basic data collection module is used to obtain elevator usage data and construction period data; A construction elevator usage graph construction module is used to extract usage features based on elevator usage data to obtain elevator usage feature data, and to construct a construction elevator usage graph based on the elevator usage feature data and construction period data to obtain construction elevator usage graph data; The construction elevator residence recommendation module is used to calculate the residence layer according to the construction elevator usage map data to obtain the construction elevator residence map data; obtain the current construction period data, and match the current construction period data and the construction elevator residence map data to obtain the construction elevator residence recommendation data; The current construction elevator residence optimization module is used to obtain the current elevator usage data, perform residence optimization based on the current elevator usage data and the construction elevator residence recommendation data, and obtain the current construction elevator residence optimization data to perform intelligent control operations on the construction elevator.
Citation Information
Patent Citations
Overload protection system of construction lifter based on big data
CN111483900A
Traffic demand predicting device of elevator
JP2010006613A
Apparatus of elevator group management control and detection method of number of persons waiting on hall of the same.
KR1019980076213A
Well construction activity graph builder
US20220335184A1
