Informationized concrete mixing plant management system

Abstract

The invention relates to the technical field of information management, in particular to an informatization concrete mixing plant management system. In the invention, through periodically distributing the state labels of the inventory remaining amount of the concrete materials and generating the time sequence, a dynamic label mapping relation can be established for inventory management of the concrete materials, and the coupling degree between the inventory state and the consumption trend is enhanced. In a process of deducing a future state change path by using a state label sequence, frequency statistics and transition probability modeling between label pairs are adopted, so that inventory change trend prediction has a data-driven evolution capability, and the adaptability and foresight of prediction are improved. In the aspect of analysis of the running state of the conveying belt, running tracks of continuous time periods are generated in a potential energy space by means of a normalized vector form of running parameter tension, belt speed and motor load, so that the running state can specifically express and monitor the jamming trend in a multi-dimensional space, and the accuracy of abnormal state recognition is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of information management, and in particular to an information-based concrete mixing plant management system. Background Art

[0002] An information-based concrete mixing plant management system refers to a system that comprehensively manages and controls a concrete mixing plant using information technology, mainly covering production scheduling, equipment monitoring, data collection and processing, etc. of the concrete mixing plant. Through an automated method, it realizes real-time monitoring of the operating status of the mixing plant equipment, automatic scheduling of production plans, management of concrete materials, and control of batching accuracy.

[0003] Existing information-based concrete mixing plant management systems mainly rely on static monitoring of production scheduling and equipment status, and can only perform ordinary scheduling based on the current inventory, making it difficult to predict the upcoming material shortage risk. In the operation monitoring of the conveyor belt, it can only judge whether there is an operation abnormality by threshold triggering, ignoring the operation trend and change trajectory within a continuous time period, which is prone to delay the recognition of the jamming trend. In terms of identifying the blocked state at the feeding port, it usually relies on sensor data at a single point in time, making it difficult to determine whether the blocked state is accidental or continuously abnormal, and is prone to misjudgment or response delay. For example, during the high-intensity production stage, if it is judged that there is a blockage only based on the increase in the motor load at a certain point in time, short-term fluctuations will be misjudged as abnormalities, resulting in unnecessary equipment shutdowns and scheduling errors, limiting the intelligent level and preventive maintenance ability of the management system. Summary of the Invention

[0004] The purpose of the present invention is to solve the shortcomings existing in the prior art, and to propose an information-based concrete mixing plant management system.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions: An information-based concrete mixing plant management system includes: The inventory status recognition module obtains the total daily consumption of concrete materials in the concrete mixing plant and calculates the remaining inventory of the corresponding concrete materials, and assigns status labels to the remaining inventory of the concrete materials according to a fixed time period to obtain a status label sequence; The state transition deduction module extracts the status label pairs in adjacent time periods in the status label sequence, predicts the future change path of the remaining inventory of the concrete materials, and obtains the inventory change prediction result; The operation potential energy mapping module obtains the operation parameters of the conveyor belt of the mixing plant and performs normalization processing, maps them to the potential energy space in chronological order, and constructs a conveyor belt potential energy distribution map; The potential energy trend recognition module divides the potential energy distribution map of the conveyor belt into multiple local neighborhoods, calculates the local potential energy density within the local neighborhoods to identify the operating state of the conveyor belt, and adjusts the operating mode of the conveyor belt in combination with the inventory change prediction result to obtain the conveyor belt regulation result; The Boolean sequence construction module, based on the conveyor belt regulation result, real-time collects the opening and closing signals of the feeding port of the mixing plant after the conveyor belt is adjusted, and constructs a Boolean state sequence during the operation of the feeding port; The feeding port abnormality analysis module, based on the collected opening and closing signals, compares the differences in the Boolean state sequences of the historical and current operating stages, judges the different material blocking states of the feeding port, and generates a warning signal for the material blocking state of the feeding port.

[0006] As a further solution of the present invention, the obtaining step of the state label sequence is specifically as follows: Obtain the total daily consumption of concrete materials in the concrete mixing plant, statistically count the corresponding concrete material consumption in each time period according to a fixed time period and compare it with the remaining amount of concrete materials in the mixing plant in the corresponding time period respectively to obtain the comparison result of concrete material consumption and remaining amount; Assign different state labels of the remaining amount of concrete materials to the comparison results of the consumption and remaining amount of concrete materials in each time period. The state labels include sufficient, critical, and out of stock. Integrate all state labels to obtain a state label sequence.

[0007] As a further solution of the present invention, the obtaining step of the inventory change prediction result is specifically as follows: Extract the state label pairs of adjacent time periods in the state label sequence, compare the state label contents of the state label pairs of adjacent time periods, judge whether the state has changed, count the transfer frequencies of the changed states in all time periods, and construct a state label transfer frequency table; Initialize the Markov chain model, input the state label transfer frequency table into the Markov chain model, and the Markov chain model predicts the transfer probability of the state label of the current remaining amount of concrete materials to the state labels of other remaining amounts of concrete materials in multiple future time periods by constructing a corresponding state label transfer probability matrix to obtain a future transfer probability prediction result; Identify the future change path of the remaining amount of concrete materials according to the future transfer probability prediction result. The future change path includes continuous out of stock of the remaining amount of concrete materials, recovery to the critical state, and gradual recovery to sufficiency, to obtain the inventory change prediction result.

[0008] As a further solution of the present invention, the obtaining step of the potential energy distribution map of the conveyor belt is specifically as follows: Obtain the operating parameters of the conveyor belt of the mixing plant and perform normalization processing. Among them, the operating parameters of the conveyor belt include the tension, belt speed, and motor load of the conveyor belt, to obtain the processed operating parameters of the conveyor belt; The processed conveyor belt operation parameters for each time period are respectively combined into corresponding vector forms, mapped to the potential energy space in chronological order. In the potential energy space, the conveyor belt operation parameters represent the conveyor belt operation state points for each time period, and a conveyor belt potential energy distribution map is constructed.

[0009] As a further solution of the present invention, the steps for obtaining the conveyor belt regulation result are specifically as follows: Based on the conveyor belt potential energy distribution map, compare it with a preset local neighborhood judgment threshold. According to the comparison result, divide the conveyor belt operation state points in the potential energy space into multiple local neighborhoods by distance to obtain a local neighborhood division result; Based on the local neighborhood division result, calculate the local potential energy density of each conveyor belt operation state point within its corresponding local neighborhood. Analyze the aggregation degree and density gradient change of each conveyor belt operation state point according to the local potential energy density, and divide the corresponding density region boundaries; Obtain the position of the conveyor belt operation parameters in the potential energy space, calculate the spatial position offset between consecutive time periods, determine the offset trajectory according to the spatial position offset, compare the offset trajectory with the density region boundary, and identify whether the conveyor belt operation state migrates towards a jamming trend according to the comparison result to obtain a conveyor belt operation state analysis result; Combine the conveyor belt operation state analysis result and the future change path of the remaining concrete material inventory in the inventory change prediction result, and adjust the operation mode of the conveyor belt with reference to the conveyor belt operation state and future inventory remaining amount to obtain a conveyor belt regulation result.

[0010] As a further solution of the present invention, the steps for obtaining the Boolean state sequence are specifically as follows: Based on the conveyor belt regulation result, collect the opening and closing signals of the feeding port of the mixing plant after the conveyor belt is adjusted in real time. Count the number of switching times of the opening and closing signals according to the opening and closing actions within a specified time. By setting an opening and closing frequency threshold, divide the time periods with the number of opening and closing switching times exceeding the opening and closing frequency threshold into abnormal closing time periods to obtain a switching times statistical result; Based on the switching times statistical result, perform a discharging response detection on the time periods in the open state. Divide the time periods when the material does not flow or the sensor continuously feedbacks a blocking signal into abnormal blocking time periods. Set Boolean state markers for abnormal closing, abnormal blocking, and corresponding normal operations, and construct a Boolean state sequence during the operation of the feeding port.

[0011] As a further solution of the present invention, the steps for obtaining the warning signal for the blocking state of the feeding port are specifically as follows: Based on all the collected opening and closing signals, the Boolean difference method is used to compare the Boolean state sequences of the historical and current operation stages, and the frequency of Boolean state transitions newly added in the Boolean state sequence of the current operation stage compared with the Boolean state sequence of the historical operation stage is calculated to identify whether the blocking phenomenon shows an abnormally increasing trend; From the statistical frequency of the Boolean state transitions, the section where the abnormal strengthening trend first appears is extracted as the initial point of the material blocking abnormality. The duration and time interval distribution of the subsequent abnormal strengthening trend are statistically analyzed to determine whether the material blocking state is an instantaneous fluctuation or a continuous abnormality. An early warning signal for the material blocking state of the feed port is generated for intervention management of the feeding link of the concrete mixing station.

[0012] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by periodically allocating the state labels of the inventory balance of concrete materials and generating a time series, a dynamic label mapping relationship can be established for the inventory management of concrete materials, and the coupling degree between the inventory status and the consumption trend can be enhanced. In the process of using the state label sequence to deduce the future state change path, the frequency statistics and transition probability modeling between the label pairs are adopted to enable the inventory change trend prediction to have data-driven evolution capabilities, thereby improving the adaptability and foresight of the prediction. In terms of the analysis of the conveyor belt operation state, with the help of the normalized vector form of the operating parameters tension, belt speed and motor load, the operation trajectory of the continuous time period is generated in the potential energy space, so that the operation state can be concretely expressed in the multidimensional space and the jamming trend can be monitored, thereby enhancing the accuracy of abnormal state identification. By calculating the potential energy density and density gradient changes in the local neighborhood, the aggregation degree of the conveyor belt operation can be deeply analyzed, and the conveyor belt operation mode can be adjusted in conjunction with the inventory change trend to achieve the coordinated control between the concrete material transportation path and the inventory dynamics. In addition, a Boolean state sequence is constructed by opening and closing signals, and the actual operation signal is mapped to the feeding process logic flow, so that the fluctuation characteristics of the feeding link can be structured. By comparing the historical and current Boolean state sequence changes, it is possible to capture the continuous enhancement characteristics in the blocking state and realize the advanced identification of the blocking trend. The overall processing logic is linked to the entire chain from concrete material consumption statistics to conveyor belt behavior regulation and feed abnormality warning. A multi-dimensional interactive feedback mechanism is constructed at the time scale, space scale and state evolution level, greatly improving the active perception and intelligent scheduling capabilities of the concrete mixing station operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 It is a system flow chart of the present invention. DETAILED DESCRIPTION

[0014] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0015] Please refer to Figure 1 , an information-based concrete mixing plant management system includes: The inventory status recognition module obtains the total daily consumption of concrete materials in the concrete mixing plant and calculates the remaining inventory of the corresponding concrete materials, and assigns status tags to the remaining inventory of the concrete materials according to a fixed time period to obtain a status tag sequence; The state transition deduction module extracts the status tag pairs in adjacent time periods in the status tag sequence, predicts the future change path of the remaining inventory of the concrete materials, and obtains the inventory change prediction result; The operating potential energy mapping module obtains the operating parameters of the conveyor belt of the mixing plant and performs normalization processing, and maps them to the potential energy space in chronological order to construct a conveyor belt potential energy distribution map; The potential energy trend recognition module divides multiple local neighborhoods of the conveyor belt potential energy distribution map, calculates the local potential energy density within the local neighborhood to identify the operating state of the conveyor belt, and adjusts the operating mode of the conveyor belt in combination with the inventory change prediction result to obtain the conveyor belt control result; The Boolean sequence construction module, based on the conveyor belt control result, real-time collects the opening and closing signals of the feeding port of the mixing plant after the conveyor belt is adjusted, and constructs a Boolean state sequence during the operation of the feeding port; The feeding port abnormality analysis module, based on the collected opening and closing signals, compares the differences between the Boolean state sequences in the historical and current operating stages, judges the different blockage states of the feeding port, and generates a warning signal for the blockage state of the feeding port.

[0016] The specific steps for obtaining the status tag sequence are as follows: Obtain the total daily consumption of concrete materials in the concrete mixing plant, count the consumption of the corresponding concrete materials in each time period according to a fixed time period and compare them with the remaining inventory of the concrete materials in the corresponding time period of the mixing plant respectively to obtain the comparison result of the concrete material consumption and the remaining amount; The total daily consumption of concrete materials is determined by statistical calculations at fixed time intervals. For example, in a concrete mixing plant, the consumption of concrete materials may be counted every hour. By recording the hourly production volume of concrete, the specific quantity of concrete materials consumed during each time period can be obtained. During this process, the material consumption data for each hour is collected and stored in a database. Next, these consumption data are compared with the remaining inventory of concrete materials during the corresponding time period. For instance, assume that at the end of a certain time period (e.g., from 9 am to 10 am), 200 tons of cement are consumed, and the cement inventory at the start of this period is 500 tons. At this time, it is necessary to compare the cement inventory with the consumption to determine whether the current inventory is sufficient to meet the subsequent production requirements. If the cement inventory is only 300 tons at this time, it indicates that replenishment may be required in the following time. After the next inventory replenishment (e.g., at 12 noon), the system will update the inventory data and conduct a new comparative analysis. Through this hourly comparison method, the consumption of concrete materials during the concrete production process can be monitored in real time, and the inventory can be ensured to be sufficient.

[0017] Assign different status labels to the remaining inventory of concrete materials for the comparison results of concrete material consumption and remaining quantity in each time period. The status labels include sufficient, critical, and out of stock. Integrate all status labels to obtain a status label sequence. According to the comparison results of concrete material consumption and remaining inventory in each time period, assign different status labels to the remaining inventory of concrete materials. Assume that the assignment of status labels is based on the following rules: Sufficient: When the remaining inventory is greater than a preset percentage of the consumption (e.g., remaining inventory ≥ 150% of the consumption), the label is "Sufficient"; Critical: When the remaining inventory is close to the consumption and close to the safety warning line (e.g., the remaining inventory is between 100% and 150%), the label is "Critical"; Out of stock: When the remaining inventory is below a certain threshold of the consumption (e.g., remaining inventory ≤ 100% of the consumption), the label is "Out of stock".

[0018] Assume that the consumption of concrete materials in each time period is (where i represents the time period number), and the remaining inventory of concrete materials in the mixing plant is , where and represent the consumption and remaining inventory in the 1st to nth time periods respectively. Use the formula: , to calculate the comparison value for the ith time period; Map the labels according to the comparison value as follows: If , the status is "Sufficient"; if , the status is "Critical"; if , the status is "Out of stock".

[0019] For example, assume that the cement consumption in the 1st hour is tons and the remaining stock is tons. Then the comparison value , and the status is "sufficient"; the consumption in the 2nd hour is tons and the remaining stock is tons. Then , and the status is "critical"; the consumption in the 3rd hour is tons and the remaining stock is tons. Then , and the status is "out of stock". The final status label sequence is: 1st hour - sufficient; 2nd hour - critical; 3rd hour - out of stock.

[0020] This result shows that by calculating the ratio of consumption to inventory for each time period, the current material status can be intuitively judged, and this status can be quantified into a standardized label, providing directly usable input conditions for subsequent data processing, status sequence integration, etc. Through the status label, the inventory situation of concrete materials in each time period can be clearly reflected, inventory problems can be identified in a timely manner, and corresponding adjustment measures can be taken.

[0021] By comparing the consumption of concrete materials within a fixed time period with the remaining stock of concrete materials in the corresponding time period segment by segment, and assigning status labels based on the comparison results, the dynamic grading and continuous marking of inventory status can be achieved, enabling inventory management to shift from static stock monitoring to dynamic status identification. By forming a status label sequence covering the entire time period, the evolution process of the relationship between concrete material consumption and remaining can be comprehensively reflected, providing data support for subsequent inventory change trend prediction, raw material replenishment scheduling, and equipment operation adjustment, and improving the real-time performance and accuracy of inventory supervision.

[0022] The specific steps for obtaining the inventory change prediction result are as follows: Extract the status label pairs of adjacent time periods in the status label sequence, compare the status label contents of the status label pairs of adjacent time periods, judge whether the status has changed, count the transfer frequencies of each changed status within all time periods, and construct a status label transfer frequency table; Based on the obtained status label sequence "1st hour - sufficient, 2nd hour - critical, 3rd hour - out of stock", assuming the status labels are continuously extracted until the 10th hour, the expanded complete sequence is: 1st hour - sufficient, 2nd hour - critical, 3rd hour - out of stock, 4th hour - out of stock, 5th hour - critical, 6th hour - critical, 7th hour - sufficient, 8th hour - sufficient, 9th hour - critical, 10th hour - out of stock. Extract the status label pairs within adjacent time periods, including "sufficient - critical", "critical - out of stock", "out of stock - out of stock", "out of stock - critical", "critical - critical", "critical - sufficient", "sufficient - sufficient", "sufficient - critical", "critical - out of stock". Compare these label pairs one by one to determine whether the adjacent status has changed. Among the above 9 label relationships, "sufficient - critical", "critical - out of stock", "out of stock - critical", "critical - sufficient", "sufficient - critical", "critical - out of stock" are the cases where the status has changed, while "out of stock - out of stock", "critical - critical", "sufficient - sufficient" are the cases where the status has not changed. Count all the state transition times and construct a state label transition frequency table. Using "sufficient, critical, out of stock" as the state set, the following frequency statistics are obtained through pairwise analysis: from "sufficient" to "critical" appears 2 times, from "critical" to "out of stock" appears 2 times, from "out of stock" to "out of stock" appears 1 time, from "out of stock" to "critical" appears 1 time, from "critical" to "critical" appears 1 time, from "critical" to "sufficient" appears 1 time, from "sufficient" to "sufficient" appears 1 time. The state transition frequencies that do not appear are all recorded as 0, and finally a complete state label transition frequency table is formed.

[0023] Initialize the Markov chain model, input the state label transition frequency table into the Markov chain model. The Markov chain model predicts the transition probability of the state label of the current concrete material inventory balance to the state label of other concrete material inventory balances within multiple future time periods by constructing a corresponding state label transition probability matrix, and obtains the future transition probability prediction result; The Markov chain model is a mathematical model used to describe the probabilistic transfer of system states between different time points. Its basic structure consists of three core elements: state set, state transition probability matrix, and initial state distribution. The state set is a finite set of all possible system states in the model. For example, in inventory management, it can include "sufficient", "critical", "out of stock", etc.; the state transition probability matrix describes the probability of transition between any two states within one time step. Each row corresponds to a current state, and each column represents the target state that may be transferred to. The sum of the elements in each row of the matrix is 1; the initial state distribution is the probability of the system being in each state when the model starts running.

[0024] Let the state set be , assuming that based on the statistically obtained state label transition frequencies, the state label transition frequency matrix is constructed as follows: the frequency from "sufficient" to "sufficient" is 1, from "sufficient" to "critical" is 2, and from "sufficient" to "out of stock" is 0; from "critical" to "sufficient" is 1, from "critical" to "critical" is 1, and from "critical" to "out of stock" is 2; from "out of stock" to "sufficient" is 0, from "out of stock" to "critical" is 1, and from "out of stock" to "out of stock" is 1.

[0025] Use the dtmc toolbox in MATLAB software to execute the Markov chain modeling process. Input the above state label transition frequency matrix as the state transition data of the Markov chain model to construct the state label transition probability matrix. , first set the state set. The state transition frequency matrix is obtained by counting the historical state sequence labels. The transition frequency of each state is obtained by traversing adjacent state label pairs in the time series and recording the occurrence times of transitions between each state.

[0026] In the Markov chain model, when currently in the first state the probability of transitioning to the second state is denoted as : ; where represents the frequency of transitioning from the first state to the second state in the state label transition frequency table; represents the total frequency of all possible transitions from the first state , that is, the total number of transitions of the first state ; is an enumeration variable used to represent all second states ; is the total number of all states in the state label transition frequency table, indicating that the state set contains three states: "sufficient", "critical", and "out of stock".

[0027] Set the state set as . According to the state label transition frequencies statistically obtained in Paragraph 1, the state label transition frequency matrix is constructed as follows: the frequency from "sufficient" to "sufficient" is 1, from "sufficient" to "critical" is 2, and from "sufficient" to "out of stock" is 0; from "critical" to "sufficient" is 1, from "critical" to "critical" is 1, and from "critical" to "out of stock" is 2; from "out of stock" to "sufficient" is 0, from "out of stock" to "critical" is 1, and from "out of stock" to "out of stock" is 1.

[0028] Taking the state "sufficient" as the initial state as an example: Statistical frequency of its transitions to other states: , , , the total transfer frequency is: , then: .

[0029] Similarly, calculate the "critical" state transition probability: , , , the total frequency is , and the corresponding probability: .

[0030] For the "material shortage" state: , , , the total frequency is , and the corresponding probability: .

[0031] Finally, a state transition probability matrix is formed: .

[0032] The current state is "material shortage", that is, the state vector at the 10th hour is: ; Using the first-order Markov state transition formula:

[0033] Among them, is the state distribution vector at the th hour, which is a row vector with a length of , representing the probability distribution of the system being in each state at the current time point. For example, means being completely in the "material shortage" state; is the predicted result of the state distribution at the th hour, indicating the probabilities of transferring to the three states of "sufficient", "critical", and "material shortage" from the current state at the next moment; is the state transition probability matrix, with a dimension of , where each row corresponds to a current state, each column represents a possible target state, and each element represents the probability of transferring from state to state ; the process of multiplying the state vector by the transition matrix represents the weighted superposition of each possible target state starting from the current distribution to obtain the state probability at the next moment.

[0034] Prediction process for the 11th hour: The initial probability vector is that the state at the 10th hour is material shortage at the same time, , multiplied by the state transition probability matrix : ;

[0035] The first item (sufficient): ; The second item (critical): ; The third item (lack of materials): .

[0036] The result is: .

[0037] It means that the probability of being in the "critical" state at the 11th hour is 0.5, the probability of being in the "lack of materials" state is 0.5, and the probability of being in the "sufficient" state is 0.

[0038] The prediction process for the 12th hour: Based on the state vector at the 11th hour , continue to use the state transition formula:

[0039] The first item (sufficient): ; The second item (critical): ; The third item (lack of materials): .

[0040] Therefore: .

[0041] It means that at the 12th hour: there is a 12.5% probability of being in the "sufficient" state; there is a 37.5% probability of being in the "critical" state; there is a 50% probability of being in the "lack of materials" state.

[0042] This distribution indicates that the inventory tight state still dominates in a short period (lack of materials is the main), but the growth of the critical state shows that the system has a slight tendency to stabilize, while the possibility of the sufficient state is still low. This state prediction can provide a basis for inventory replenishment scheduling and prompt the need to organize the replenishment action before the 12th hour.

[0043] The calculation of the state transition probability matrix is obtained by counting the transition frequencies between various states in the historical state label sequence. This process can quantify the discrete historical state change behavior into a systematic probability relationship. The calculation logic of the first-order Markov state transition formula is based on the state distribution vector at the current moment and the state transition probability matrix multiplied to obtain the state prediction distribution , indicating that the next state of the system depends only on the current state, and is calculated by superimposing the probabilities of transferring to each target state according to the current state. Through this process, the model can not only predict the likelihood of being in different states at any future time period, but also reveal the trend of system state changes, providing quantitative support for decision-making such as inventory control, material replenishment, or risk warning.

[0044] Identify the future change path of the inventory balance of concrete materials based on the future transfer probability prediction results. The future change path includes continuous shortage of the inventory balance of concrete materials, recovery to the critical state, and gradual recovery to sufficiency, and obtain the inventory change prediction results; First, according to the state transition vector generated by the state "shortage" at the 10th hour , it can be known that after entering the 11th hour, there are two equally probable evolution directions for the inventory state of concrete materials, that is, the probability of maintaining the "shortage" state is 50%, and the probability of slowly recovering to the "critical" state is also 50%. Under this prediction condition, the system is in an unstable fluctuation state. The coexistence of shortage and critical means that the inventory tension risk still exists but there are already some signs of recovery. Continuing to predict to the 12th hour, we get , the state transition probability distribution further changes. Among them, the "shortage" state still has the largest proportion, maintaining a proportion of 50%, but the "critical" state rises to 37.5%, indicating that the inventory of concrete materials is in a continuous recovery process. Although the "sufficient" state only accounts for 12.5%, it has already appeared. Looking at the evolution path from the shortage state, the most likely paths are "shortage → shortage → shortage" or "shortage → critical → shortage", and there is also a certain probability of "shortage → critical → critical" or "shortage → critical → sufficient". If the future state path enters the continuous evolution of "shortage → shortage → shortage", it means that the replenishment measure has not intervened and the risk will continue to remain; if it enters the path of "shortage → critical → sufficient", it indicates that the inventory will gradually return to stability. The current prediction results do not show a strong inventory recovery signal. Therefore, the change path shows a trend of "oscillatory recovery", and it is necessary to further extend the prediction period or increase the transfer probability of intermediate states to achieve the return of the inventory safety state, so as to finally obtain the trend prediction results of inventory changes.

[0045] By extracting the state change relationship between adjacent time periods in the state label sequence and counting the state transition frequencies, the evolution law of the inventory state of concrete materials can be established. With the help of the Markov chain model to construct the state transition probability matrix, the quantitative prediction of the inventory state change trend in multiple future time periods can be realized, enabling the concrete material management to have the ability of trend identification and path evolution judgment. By identifying the future change path of the inventory balance, trends such as shortage and recovery can be perceived in advance, enhancing the forward-looking and decision-making support capabilities of inventory management, and avoiding sudden interference to production caused by shortages of concrete materials.

[0046] The steps for obtaining the potential energy distribution diagram of the conveyor belt are specifically as follows: Obtain the operating parameters of the conveyor belt in the mixing plant and perform normalization processing. Among them, the operating parameters of the conveyor belt include the tension, belt speed, and motor load of the conveyor belt, and obtain the processed operating parameters of the conveyor belt; First, install sensors of corresponding types at key parts of the conveyor system in the mixing plant to achieve real-time acquisition of the operating parameters of the conveyor belt. The conveyor belt tension is measured by installing strain gauge type tension sensors on both sides of the conveyor belt tensioning device. This type of sensor can output an electrical signal proportional to the magnitude of the tension. The acquisition of the belt speed depends on the rotary encoder installed at the end of the conveyor belt drive drum or return drum. The encoder indirectly reflects the belt line speed by recording the number of rotation pulses per unit time. The motor load data is jointly acquired by current sensors and voltage sensors installed on the input cable path of the main drive motor. This group of sensors can real-time read the current intensity and voltage value during the operation of the motor and output a data frame. All three types of sensors are connected to the data acquisition module, and the acquisition period is set to once every 5 seconds. The acquired data is uniformly imported into the data management platform for normalization processing. The normalization process uses the pandas and sklearn toolkits in Python. By calling the standardized data preprocessing module, the tension, belt speed, and motor load data are independently converted into standardized values within the range of 0 to 1. The normalization process is performed separately for each type of parameter, and the standardized results will be used as the data input for subsequent operating state analysis.

[0047] Combine the processed operating parameters of the conveyor belt in each time period into corresponding vector forms respectively, and map them to the potential energy space in chronological order. In the potential energy space, the operating parameters of the conveyor belt represent the operating state points of the conveyor belt in each time period, and construct the potential energy distribution diagram of the conveyor belt; Combine the processed operating parameters of the conveyor belt in each time period into corresponding vector forms respectively. The normalized tension, belt speed, and motor load data within each time period are combined into a three-dimensional column vector according to the unified time stamp. The vector form is expressed as: ; where represents the operating state vector of the conveyor belt in the th time period of the processed operating parameters of the conveyor belt, which is dimensionless, , , are respectively the normalized values of the tension, belt speed, and motor load of the processed operating parameters of the conveyor belt in the th time period. This three-dimensional vector is used to accurately describe the multi-dimensional state characteristics of the conveyor belt operation at this time point. Then, the vectors of each time period Arrange them in chronological order and map them to the potential energy space. In the potential energy space: Tension represents the strength of the tensile state of the conveyor belt, reflecting the stress level of the belt tensioning system. Excessive tension may indicate overload, and low tension may indicate belt slippage or slack; Belt speed represents the linear movement speed of the conveyor belt, representing the material passing rate per unit time. Changes in belt speed can reveal whether the operation is stable; Motor load represents the energy consumption state of the drive system, which is usually closely related to transmission resistance and load size. Fluctuations in motor load may reveal mechanical abnormalities or fluctuating operating conditions. These three-dimensional parameters correspond to the X, Y, and Z axes of the spatial coordinate system in the potential energy space: Tension is the X-axis component, belt speed is the Y-axis component, and motor load is the Z-axis component. The operating state points for each time period are uniquely located by this three-dimensional vector in the three-dimensional space, forming a state trajectory that evolves continuously over time. Further, by performing three-dimensional space interpolation and density analysis on the set of all operating state points, a potential energy distribution map of the conveyor belt operation state can be constructed, which is used to visualize the operation stability, identify high-energy consumption or high-stress concentration areas, and complete the temporal-spatial expression of the conveying state. The construction and visualization of this map can be achieved using the plot3 and surf functions in MATLAB or the Axes3D.plot and plot_surface functions in the matplotlib library of Python. If higher interactivity and display accuracy are required, it can also be achieved using the scatter3d and surface modules of Plotly. During the plotting process, different color gradients can be assigned to the state points of different time periods to enhance the recognition of the operation evolution rhythm, and finally a complete potential energy distribution map of the conveyor belt is generated.

[0048] By normalizing multiple operating parameters such as the tension, belt speed, and motor load of the conveyor belt and mapping them to the potential energy space in vector form, the operating state of the conveyor belt in each time period can be visually expressed as potential energy state points, thus constructing a complete operating trajectory map of the conveyor belt. This method can capture the internal relationships between operating parameters, reveal the evolution trend of the equipment operating state, improve the refined analysis ability of the dynamic behavior of the conveyor belt, and provide multi-dimensional basis for anomaly detection and control strategies.

[0049] The specific steps for obtaining the conveyor belt control result are as follows: Based on the potential energy distribution map of the conveyor belt, compare it with the preset local neighborhood judgment threshold, and divide the conveyor belt operating state points in the potential energy space into multiple local neighborhoods according to the distance based on the comparison result to obtain the local neighborhood division result; Based on the operating state points mapped to the potential energy space, using the distance metric between points in a three-dimensional Euclidean space, with each state point as the center, a spatial radius threshold is selected, and all neighboring state points within this radius that are no more than the radius threshold away from the center point are extracted to form the local neighborhood set corresponding to this point. The distance calculation uses the Euclidean distance formula in a three-dimensional space. Taking the ternary coordinates composed of the tension, belt speed, and motor load normalization values of each point as the input, the spatial distance between any two points is calculated in batches, and then neighborhood clustering is performed on each point. In the process, the KDTree structure in the scipy library in Python is used to achieve fast neighbor search, forming a set of neighborhood points corresponding to each state point, and recording the number of state points included in each neighborhood group. After all state points have completed neighborhood division, a local neighborhood set automatically grouped according to the spatial position density is obtained.

[0050] Calculate the three-dimensional Euclidean distance between any two conveyor belt operating state points in the potential energy space. The formula is as follows:

[0051] Where, represents the Euclidean distance (dimensionless) between the th conveyor belt operating state point and the th conveyor belt operating state point in the potential energy distribution map of the conveyor belt, that is, the magnitude of the position difference in the three-dimensional system of tension - belt speed - load; are the respective numbered indices of the conveyor belt operating state points in the potential energy distribution map of the conveyor belt, representing the operating state points sampled at different time periods; are respectively and corresponding tension normalization values; are respectively and corresponding belt speed normalization values; are respectively and corresponding motor load normalization values.

[0052] Based on vector form representation: , assuming the first state point , and the sampling time period is 09:00, then ; specifically, the tension at 09:00 is 0.60 (after normalization), the belt speed is 0.80, and the motor load is 0.45. The second state point , and the sampling time period is 09:05, then ; specifically, the tension at 09:05 is 0.70 (after normalization), the belt speed is 0.65, and the motor load is 0.50. The third state point The sampling time period is 09:10, then ; specifically, the tension at 09:10 is 0.62 (after normalization), the belt speed is 0.78, and the motor load is 0.40

[0053] Calculate the Euclidean distance between the first state point 1 and the second state point 2 in the potential energy space: ; Calculate the Euclidean distance between the first state point 1 and the third state point 3 in the potential energy space: ; Calculate the Euclidean distance between the second state point 2 and the third state point 3 in the potential energy space: ; The neighborhood judgment threshold can be selected according to the distance distribution curve of all point pairs. The minimum 10% value can be used as the reference threshold for the adjacent region. Alternatively, multiple sets of normal and abnormal data can be tested, and the maximum spatial span can be observed in the obvious clustering region, and the upper bound of the maximum span can be used as the reference threshold for the adjacent region. Assume that the local neighborhood threshold is represented. When the three-dimensional distance between two state points is less than this value, it can be considered that their operating states are similar and belong to the same local neighborhood → belong to the same neighborhood; → belong to the same neighborhood; → belong to the same neighborhood. The results show that the pairwise distances between the three state points all meet the neighborhood conditions, indicating that the variation range of their operating characteristics in the three dimensions of tension, speed, and load is small, indicating that the conveyor belt operates stably during this period and the parameter performance is consistent

[0054] Through the calculation of the normalized parameters and Euclidean distances of specific operating state points, it can intuitively reflect whether the operating state differences of the conveyor belt at different times are significant. If the pairwise distances between multiple points are all within the threshold, it can be considered that these points belong to the same operating mode in space, and subsequent density calculations and trend judgments can be performed based on this neighborhood

[0055] Based on the local neighborhood division results, calculate the local potential energy density of each conveyor belt operating state point in its corresponding local neighborhood, analyze the aggregation degree and density gradient change of each conveyor belt operating state point according to the local potential energy density, and divide the corresponding density region boundary; Calculate the th conveyor belt operating state point in its corresponding local neighborhood The formula for the local potential energy density (dimensionless) is as follows:

[0056] Among them, is the summation index range, indicating all other th conveyor belt operating state points within the local neighborhood corresponding to the th conveyor belt operating state point, excluding itself; th conveyor belt operating state point; is the natural exponential function, used for exponential decay processing of the square of the distance; represents the Euclidean distance in the potential energy space between the th conveyor belt operating state point and the th conveyor belt operating state point in the potential energy distribution diagram of the conveyor belt; is the kernel width parameter for regulating the decay rate of the exponential function in the calculation of local potential energy density, controlling the amplitude of density decay. The value is determined by the structural change range of the density region. For example, it can be determined by analyzing the median of the distance distribution between multiple batches of sample point pairs. For example, it can be initially screened by using the statistical standard deviation method of batch operation data, and then the appropriate value can be selected through repeated verification of neighborhood aggregation comparison experiments.

[0057] Given the first state point and the second state point in the potential energy space, the Euclidean distance: ; the first state point and the third state point in the potential energy space, the Euclidean distance: ; the second state point and the third state point in the potential energy space, the Euclidean distance: , Calculate the density of the state point in the neighborhood : ; Calculate the density of the state point in the neighborhood : ; Calculate the density of the state point in the neighborhood : .

[0058] Given that the local potential energy densities of each conveyor belt operating state point are: for state point 1 (09:00) is , for state point 2 (09:05) is , and for state point 3 (09:10) is .

[0059] By comparing the density values, we can conclude that: state point 3 has the highest density and is the most concentrated point in the local neighborhood, which can be regarded as a dense center; state point 1 has the second highest density and is located in a relatively concentrated area, close to the density peak point; state point 2 has the lowest density, is far away from the aggregation core, and is in a sparse edge area.

[0060] Calculate the density difference between any two state points as the local density gradient value:

[0061]

[0062]

[0063] The results show that the density drops sharply from state point 3 to state point 2, and there is an obvious density gradient direction; the density of state point 1 and state point 3 is similar, belonging to the same high-density area or slowly changing area; from state point 1 to state point 2, there is also an obvious downward trend. Therefore, state point 2 is a candidate for the boundary point, state point 3 is the gathering center, and state point 1 is a transition or secondary center point.

[0064] Use the average density and standard deviation of the current three points as the basis for boundary classification: Average density : ; Standard Deviation (approximate): .

[0065] According to the density distribution analysis method commonly used in statistics, the state point density is divided into the following three categories: High density area: local potential energy density value ; Medium density area: ,Right now ; Low density areas: .

[0066] result : Falling into the medium-density area, tending to high values, it is judged to be inside the cluster; : They belong to the medium-density area, and the degree of aggregation is close to point 3; : Below the low-density critical value, it is determined as a boundary edge point. Among them, the medium-density area is composed of multiple adjacent high-density state points, which is a local clustering area; the low-density area is an area with less sparse distribution points and has boundary characteristics; the regional boundary is drawn from the medium-density to the low-density gradient descent direction to form a segmentation surface.

[0067] The calculation logic of the local potential energy density is based on the relative distance relationship between the state point and other state points in its neighborhood in the potential energy space. By exponentially weighting and summing the Euclidean distances between each state point and other points in the neighborhood, the aggregation degree value of the state point in the local space is obtained. The higher the density, the more state points are adjacent to and concentrated around the point; the lower the density, the fewer adjacent points or the sparser the distribution around the point. This calculation logic not only reflects the local aggregation characteristics of a single state point in space but also provides a quantitative basis for subsequent density gradient analysis and boundary division. In the potential energy space, the differences in local potential energy density values can reveal the structural relationships between the aggregation region, the transition region, and the boundary region, thus enabling the effective identification of the clustering characteristics and evolution trends of the operating state.

[0068] Obtain the position of the conveyor belt operating parameters in the potential energy space, calculate the spatial position offset between consecutive time periods, determine the offset trajectory based on the spatial position offset, compare the offset trajectory with the density region boundary, and identify whether the conveyor belt operating state migrates towards a jamming trend according to the comparison result; After obtaining the position of the conveyor belt operation parameters in the potential energy space, according to the three-dimensional vector composed of the normalized tension, belt speed, and motor load corresponding to each time period, the operation state points of each time period are mapped to the potential energy space as three-dimensional coordinate points. For example, the state point at 09:00 is 0.60, 0.80, 0.45; at 09:05 it is 0.70, 0.65, 0.50; at 09:10 it is 0.62, 0.78, 0.40; these points form a spatial trajectory within consecutive time periods. Subsequently, based on the vector coordinate differences between every two adjacent time points, the Euclidean distance is used to calculate the offset. For example, the offset from 09:00 to 09:05 is 0.187, and from 09:05 to 09:10 is 0.182, thus establishing a complete sequence of trajectory segments. Next, the trajectory line is spatially overlapped with the boundary of the previously defined density region to determine the region where each state point is located in the potential energy space. For example, the densities at 09:00 and 09:10 are 1.339 and 1.359 respectively, both in the medium density region, while the density at the 09:05 point is 0.854, falling into the low density region. Therefore, the trajectory line starts in the medium density region at 09:00, then offsets to 09:05 when it crosses the medium density boundary and enters the low density region, and then offsets to 09:10 when it returns to the medium density region. Through this comparison process, it can be confirmed that there is an "out-of-bounds" behavior at the 09:05 state, that is, it offsets to the low density extension. This out-of-bounds phenomenon indicates that there are significant fluctuations in the conveyor belt operation parameters during this period, including abnormal combinations such as increased tension, decreased belt speed, and increased motor load, resulting in the operation state moving away from the stable aggregation region, characterizing the potential jamming characteristics or unstable operation signs of the conveying system during this time period. Subsequently, the trajectory rebounding back to the medium density region can be regarded as the self-recovery performance of the system, but the "out-of-bounds" is an early sign of the critical jamming trend and needs to be monitored and identified.

[0069] Combining the future change path of the inventory balance of concrete materials in the analysis result of the conveyor belt operation state and the prediction result of inventory changes, referring to the conveyor belt operation state and the future inventory balance, adjust the operation mode of the conveyor belt to obtain the conveyor belt regulation result; Combined with the prediction results of the future change path of the concrete material inventory surplus, and based on the actual deviation trajectory of the conveyor belt operation status, the operation mode of the conveyor belt is adaptively adjusted under various scenarios. For example, when the operation status is in a stable range, that is, the potential energy space trajectory does not deviate from the medium-high density area, but the prediction shows that the future inventory surplus will continue to decrease and approach the "material shortage" state. At this time, the conveyor belt operation mode can be adjusted to the deceleration mode, reducing the conveyor belt speed or intermittent operation to avoid idling operation when the material is insufficient. Another example is that the current operation status of the conveyor belt shows frequent deviation and has appeared continuous "out-of-bounds" phenomena, indicating an abnormal operation trend. However, the concrete material inventory remains "sufficient" in the next few time periods. At this time, a stable-load and slow-conveyance mode should be adopted, appropriately reducing the belt speed and increasing the operation monitoring interval to reduce the high-load transmission under abnormal working conditions. Another example is that the current conveyor belt operation status is stable and the inventory will also be sufficient in the future. Then the current operation mode can be maintained or the belt speed can be increased according to the plan to improve the material turnover efficiency. On the contrary, if the conveyor belt operation status is unstable and the inventory prediction shows a downward trend, a dual protection mode should be adopted, combining speed limit operation and intermittent control methods to avoid the dual risks of transmission failures and resource waste. Through the responsive adjustment of the above various combined situations, the coordinated linkage between the transmission system and the inventory change trend can be achieved, ensuring the safety of the material conveying process and the rationality of resource utilization.

[0070] By dividing the local neighborhood of the operation status points in the potential energy distribution map according to the distance and calculating the local potential energy density, the aggregation characteristics and density gradient changes of the operation status points can be identified, the potential energy concentration area and boundary can be clarified, and the spatial recognition ability of the conveyor belt operation stability can be enhanced. Further combined with the position offset under the time series, a continuous offset trajectory is constructed and compared with the density boundary to accurately judge whether the operation status evolves towards the jamming trend. Based on this judgment and combined with the inventory change path, the conveyor belt operation mode is adjusted to achieve the collaborative optimization of the conveyor belt behavior and the inventory status, improve the conveying efficiency and effectively avoid abnormal operation.

[0071] The specific steps for obtaining the Boolean state sequence are as follows: Based on the conveyor belt regulation results, the opening and closing signals of the feeding port of the mixing plant after the conveyor belt adjustment are collected in real time. The opening and closing signals are statistically analyzed according to the number of switching times of the opening and closing actions within the specified time. By setting the opening and closing frequency threshold, the time periods with the number of opening and closing switching times exceeding the opening and closing frequency threshold are divided into abnormal closing time periods to obtain the switching times statistical result; During each time period, the opening and closing states of the feeding port of the mixing plant are collected in real time. The collection method obtains two state change signals of opening and closing through a Hall sensor or a reed switch device installed on the feeding mechanism, and records the number of switches within the set time period to form an opening and closing action statistical value. At the same time, an opening and closing frequency threshold is set as the judgment benchmark. The setting basis of this threshold is the maximum opening and closing action frequency allowed by the mechanical structure of the feeding port. This value is usually given by the equipment manufacturer or can be obtained through long-term on-site operation monitoring to obtain empirical data. For example, under normal operating conditions, the opening and closing times of 30 time periods are continuously collected daily, and the average frequency and variance range are statistically analyzed. After excluding special working condition data, the highest value in the stable frequency band is extracted as the reference value and set as the opening and closing frequency threshold. For example, if the average opening and closing frequency within a 60-second statistical cycle is 3.2 times and the maximum stable value is 4.5 times, the opening and closing frequency threshold can be set to 5 times per minute. When the number of opening and closing switches in a certain time period exceeds this threshold, it indicates that there is a high-frequency opening and closing phenomenon in this time period, which is judged as a frequent switching action triggered by equipment abnormality, and this time period is classified as an abnormal closing state.

[0072] Based on the statistical result of the number of switches, the discharging response detection is carried out for the time periods in the open state. The time periods when the material does not flow or the sensor continuously feeds back the material blocking signal are classified as abnormal blocking time periods. Boolean state marks for abnormal closing, abnormal blocking, and corresponding normal operation are set, and a Boolean state sequence during the operation of the feeding port is constructed. For the time periods in the open state, the discharging response detection is carried out. First, the time periods when the feeding port remains in the "open" state during the statistical cycle are screened out. For example, if the opening and closing signals do not switch during the time period from 09:02 to 09:03, it is determined that this cycle is in a continuous open state. Subsequently, a material flow sensor (such as an infrared photoelectric switch or a weighing induction plate) located below the feeding port is called to record whether the material flows. If the continuous detection results of the sensor are zero flow or no change value throughout the time period, and at the same time, it is checked whether the feedback signal of the pressure sensor on the side wall of the feeding port continuously remains in a high value state, indicating that the material blockage has not been released, then this time period is identified as the material not being discharged during the open state. According to the feedback of the dual sensors, this time period is determined as an abnormal blocking state. This identification method avoids misjudgment caused by a single signal error. Finally, the abnormal blocking time period is set to the Boolean value "0", which is marked as an abnormal state in the same way as the aforementioned abnormal closing state, while other continuously open time periods with normal discharging flow detected are marked as the Boolean value "1", indicating that the feeding port is in a normal working state during this time period. After integrating the state marks of all time periods in chronological order, a complete Boolean state sequence during the operation of the feeding port is formed.

[0073] By collecting and statistically analyzing the switching frequency of the opening and closing signals of the feed inlet in real time, abnormal closing events can be efficiently identified, and the abnormal state can be accurately marked by setting the opening and closing frequency threshold. At the same time, the material flow under the open state is detected. If a blockage phenomenon is found, it can be promptly identified and marked as an abnormal blockage. Converting these state information into a Boolean state sequence can clearly present the operating conditions during the feeding process, improve the monitoring and early warning capabilities for feed inlet abnormalities, and contribute to early intervention and reduction of production interruption risks.

[0074] The steps for obtaining the warning signal of the blocked state of the feed inlet are specifically as follows: Based on all the collected opening and closing signals, the Boolean difference method is used to compare the differences between the Boolean state sequences in the historical and current operating stages, and calculate the newly added Boolean state transition frequency in the Boolean state sequence of the current operating stage compared with that of the historical operating stage, which is used to identify whether the blockage phenomenon shows an abnormal increasing trend; Let the Boolean state sequence in the historical operating stage be , and the Boolean state sequence in the current operating stage be , where represents the opening and closing state of the feed inlet at each time point in the historical time period (1 means open, 0 means closed), which is recorded by on-site sensors or exported from historical data. represents the opening and closing state of the feed inlet at each time point in the current time period (1 means open, 0 means closed), which is generated from the signals collected by sensors in real time. Then the Boolean state difference sequence is expressed as:

[0075] Among them, is the Boolean state value of the th time period in the Boolean state sequence of the historical operating stage, and the value is 0 or 1, which is provided by the historical operating data; is the Boolean state value of the th time period in the Boolean state sequence of the current operating stage, and the value is 0 or 1, which is provided by sensors in real time; is the total number of Boolean state transitions between the Boolean state sequence of the current operating stage and that of the historical operating stage, which reflects the change frequency of the state and is used to identify whether the blockage phenomenon intensifies; is the index of the time period, indicating that the Boolean states in each time period are compared, from 1 to ; is the total number of time periods, indicating the number of time periods to be compared. For example, if it is counted once per minute, then is the total number of time periods.

[0076] Suppose there are the following historical Boolean state sequence and current Boolean state sequence: Historical state sequence : [1, 1, 0, 1, 0, 1], current state sequence : [1, 0, 0, 1, 1, 0], calculate the Boolean state difference sequence : , , , , , .

[0077] Result: .

[0078] This result indicates that: compared with the historical state, the Boolean state has changed 3 times. These changes may be caused by abnormal switching of the feed inlet or blockage problems, and it is worth further analyzing whether there is a blockage phenomenon or other problems.

[0079] By comparing the Boolean state sequences of the historical and current stages for each time period, calculate the difference in the Boolean state within each time period, that is, the difference between the historical state and the current state in the corresponding time period. Then, by summing up the differences of all time periods, the total number of Boolean state transitions is obtained. This value reflects the change frequency of the Boolean state in the current operating stage compared with the historical stage. The larger the value, the more frequent the abnormal state changes. Through this calculation, the state changes in the current operating stage can be quantified, so as to identify whether the blockage phenomenon shows an abnormal increasing trend. When the calculated Boolean state difference increases significantly, it indicates that there may be abnormal blockage behavior in the system. Especially when the frequency of new Boolean state transitions is high, it means that the blockage problem may be intensifying. Therefore, the calculation process of this formula helps to identify potential blockage trends in advance by monitoring the changes in the Boolean state.

[0080] Extract the section where the abnormal increasing trend first appears from the statistically counted Boolean state transition frequencies as the initial point of the blockage anomaly, count the duration and time interval distribution of subsequent abnormal increasing trends, judge whether the blockage state belongs to instantaneous fluctuation or continuous anomaly, and generate a warning signal for the blockage state of the feed inlet for the intervention management of the feeding link of the concrete mixing plant; First, by analyzing the Boolean state difference sequence, when it is detected that the frequency of state transitions exceeds a set threshold within several consecutive time periods and this frequency continues to increase on the time axis, it is determined as the initial point of the abnormal trend of material blockage. For example, if the transition frequency of the Boolean difference sequence in a certain time period exceeds the normal fluctuation range of historical data and continues to increase, it can be determined that this time period is the initial point of material blockage abnormality. Next, the duration and time interval distribution of the subsequent abnormal enhancement trend are statistically analyzed. By continuously monitoring the next time period, if the transition frequency continues to increase and the interval time gradually shortens, it indicates that the material blockage phenomenon may evolve from a temporary fluctuation to a continuous abnormality. Otherwise, it may only be an instantaneous fluctuation. The specific calculation of the duration and time interval can be obtained by comparing the start and end times of the abnormal time period, getting the duration of each abnormality and the interval time between abnormalities. If the abnormal duration is long and the time interval is short, this trend is very likely to be a continuous blockage state. If the abnormal fluctuation only appears in a short time period and the time interval is long, it can be classified as an instantaneous fluctuation. According to the analysis results and combined with the warning value preset in the historical data, a warning signal for the material blockage state at the feed inlet is generated. When the material blockage state shows a continuous abnormality, the system will trigger a warning and send an alarm to the operator, indicating that intervention management may be needed, and appropriate measures such as adjusting the opening and closing mode of the feed inlet and reducing the conveyor belt speed are taken to avoid the impact of the material blockage phenomenon on the production process and ensure the normal operation and efficient operation of the mixing plant.

[0081] By comparing the differences between the historical and current Boolean state sequences through the Boolean difference method, it is possible to accurately identify the change trend of the material blockage phenomenon at the feed inlet and calculate the frequency of new state transitions, so as to discover whether the material blockage problem shows an abnormal enhancement trend. By extracting the initial point of the abnormal trend and analyzing its duration and time interval, it can be judged whether the material blockage state is an occasional instantaneous fluctuation or a continuous abnormality, and a material blockage warning signal is generated in a timely manner to help effectively intervene in the feeding link of the concrete mixing plant and reduce production delays or equipment damage caused by material blockage.

[0082] The above is only a preferred embodiment of the present invention and is not intended to limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An information-based concrete mixing station management system, characterized in that: The system includes: The inventory status recognition module obtains the total daily concrete material consumption of the concrete mixing station and counts the corresponding concrete material inventory balance, assigns the status label of the concrete material inventory balance according to a fixed time period, and obtains the status label sequence; The state transition deduction module extracts the state label pairs of adjacent time periods in the state label sequence, predicts the future change path of the concrete material inventory balance, and obtains the inventory change prediction result; The operation potential energy mapping module obtains the operation parameters of the mixing station conveyor belt and normalizes them, maps them to the potential energy space in chronological order, and constructs the conveyor belt potential energy distribution map; The potential energy trend identification module divides the potential energy distribution map of the conveyor belt into multiple local neighborhoods, calculates the local potential energy density in the local neighborhood to identify the operating status of the conveyor belt, and adjusts the operating mode of the conveyor belt in combination with the inventory change prediction results to obtain the conveyor belt control results; The Boolean sequence construction module collects the opening and closing signals of the mixing station feed port in real time based on the conveyor belt control results after the conveyor belt is adjusted, and constructs the Boolean state sequence during the operation of the feed port; The feed inlet abnormality analysis module compares the Boolean state sequences of the history and current operation stages based on the collected opening and closing signals, determines the different blocking states of the feed inlet, and generates a feed inlet blocking state warning signal.

2. The information-based concrete mixing station management system according to claim 1 is characterized in that: The steps of obtaining the state tag sequence are specifically as follows: Obtain the total daily concrete material consumption of the concrete mixing station, count the corresponding concrete material consumption in each time period according to a fixed time period, and compare it with the concrete material inventory balance of the mixing station in the corresponding time period to obtain the comparison results of concrete material consumption and balance; Different status labels of concrete material inventory surplus are assigned to the concrete material consumption and surplus comparison results in each time period. The status labels include sufficient, critical and shortage. All status labels are integrated to obtain a status label sequence.

3. The information-based concrete mixing station management system according to claim 1 is characterized in that: The steps for obtaining the inventory change forecast result are specifically as follows: Extracting state label pairs of adjacent time periods in the state label sequence, comparing the state label contents of the state label pairs of adjacent time periods, determining whether the state has changed, counting the transition frequencies of the changed states in all time periods, and constructing a state label transition frequency table; Initializing the Markov chain model, inputting the state label transfer frequency table into the Markov chain model, and the Markov chain model predicts the transfer probability of the state label of the current concrete material inventory surplus to the state label of other concrete material inventory surpluses in multiple future time periods by constructing a corresponding state label transfer probability matrix, and obtains a future transfer probability prediction result; According to the future transfer probability prediction result, the future change path of the concrete material inventory balance is identified, and the future change path includes the continuous shortage of concrete material inventory balance, the rebound to the critical state, and the gradual recovery to sufficiency, so as to obtain the inventory change prediction result.

4. The information-based concrete mixing station management system according to claim 1 is characterized in that: The steps for obtaining the potential energy distribution diagram of the conveyor belt are specifically as follows: Acquire and normalize the conveyor belt operation parameters of the mixing station, wherein the conveyor belt operation parameters include the tension, belt speed and motor load of the conveyor belt, and obtain the processed conveyor belt operation parameters; The processed conveyor belt operation parameters of each time period are respectively combined into corresponding vector forms and mapped to the potential energy space in chronological order. In the potential energy space, the conveyor belt operation parameters represent the conveyor belt operation state points in each time period, and a conveyor belt potential energy distribution diagram is constructed.

5. The information-based concrete mixing station management system according to claim 4 is characterized in that: The steps for obtaining the conveyor belt control result are specifically as follows: Based on the conveyor belt potential energy distribution map, the conveyor belt potential energy distribution map is compared with a preset local neighborhood judgment threshold value, and according to the comparison result, the conveyor belt running state point is divided into multiple local neighborhoods according to the distance in the potential energy space to obtain the local neighborhood division result; Based on the local neighborhood division result, the local potential energy density of each conveyor belt operation status point in its corresponding local neighborhood is calculated, the aggregation degree and density gradient change of each conveyor belt operation status point are analyzed according to the local potential energy density, and the corresponding density area boundary is divided; Obtain the position of the conveyor belt operation parameters in the potential energy space, calculate the spatial position offset between consecutive time periods, determine the offset trajectory according to the spatial position offset, compare the offset trajectory with the density area boundary, and identify whether the conveyor belt operation state is migrating towards a blocking trend according to the comparison result, and obtain the conveyor belt operation state analysis result; Combined with the conveyor belt operation status analysis result and the future change path of the concrete material inventory balance in the inventory change prediction result, the conveyor belt operation mode is adjusted with reference to the conveyor belt operation status and the future inventory balance to obtain the conveyor belt control result.

6. The information-based concrete mixing station management system according to claim 1 is characterized in that: The steps for obtaining the Boolean state sequence are specifically as follows: Based on the conveyor belt control result, the opening and closing signals of the mixing station feed port after the conveyor belt is adjusted are collected in real time, and the opening and closing signals are counted according to the number of switching of the opening and closing actions within a specified time. By setting the opening and closing frequency threshold, the time period when the number of opening and closing switching exceeds the opening and closing frequency threshold is divided into an abnormal closing time period, and the switching number statistical result is obtained; Based on the statistical results of the switching times, the discharge response detection is performed on the time period that remains in the open state, the time period when the material does not flow or the sensor continuously feeds back a blocking signal is divided into an abnormal blocking time period, and the Boolean state marks of abnormal closure, abnormal blocking and corresponding normal operation are set to construct a Boolean state sequence during the operation of the feed port.

7. The information-based concrete mixing station management system according to claim 1 is characterized in that: The steps for obtaining the feed port blocking state warning signal are specifically as follows: Based on all the collected opening and closing signals, the Boolean difference method is used to compare the Boolean state sequences of the historical and current operation stages, and the frequency of Boolean state transitions newly added in the Boolean state sequence of the current operation stage compared with the Boolean state sequence of the historical operation stage is calculated to identify whether the blocking phenomenon shows an abnormally increasing trend; From the statistical frequency of the Boolean state transitions, the section where the abnormal strengthening trend first appears is extracted as the initial point of the material blocking abnormality. The duration and time interval distribution of the subsequent abnormal strengthening trend are statistically analyzed to determine whether the material blocking state is an instantaneous fluctuation or a continuous abnormality. An early warning signal for the material blocking state of the feed port is generated for intervention management of the feeding link of the concrete mixing station.

Images