Safety management system and method for automatic control of stacker

Abstract

The invention discloses a safety management system and method for automatic control of a stacking machine, and relates to the technical field of automatic control. Through vertical layered monitoring and multi-dimensional vibration data analysis, a goods shelf is divided into independent monitoring areas according to layer numbers, and the vibration frequency and amplitude of the stacking machine on the same layer are integrated for transverse comparison; structural stress change trends at different heights are accurately identified, and misjudgment caused by single-dimension data is avoided. Historical maintenance data and a real-time vibration prediction model are integrated, a time sequence prediction model is combined to dynamically evaluate the future vibration trend of equipment, a maintenance risk value and a vibration risk value are comprehensively calculated, a traditional lagging maintenance mode is broken through, and data support is provided for active maintenance decision making. Potential risk equipment clusters are identified by using a cosine similarity matching technology, hidden dangers of similar equipment are positioned in advance in combination with a prediction and early warning mechanism, upgrading from single-machine risk management and control to cluster risk monitoring is realized, and the accuracy and response speed of maintenance resource allocation are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic control, and specifically to an automatic control safety management system and method for a stacker crane. Background Art

[0002] The stacker crane accurately locates the target storage location along the shelf aisle through the horizontal and vertical movement systems, and completes the storage and retrieval operations of goods with the help of a telescopic fork, realizing the efficient turnover of goods in the automated stereoscopic warehouse. In the existing warehousing and logistics system, the monitoring of the stacker crane mostly relies on single-dimensional data, such as running time, load weight, etc. Due to the lack of comprehensive consideration and analysis of multi-dimensional parameters, managers cannot detect potential equipment failure risks in a timely manner. This makes the equipment maintenance work relatively lagging, not only increasing the possibility of downtime, but also increasing the operation and maintenance costs. Summary of the Invention

[0003] The purpose of the present invention is to provide an automatic control safety management system and method for a stacker crane to solve the problems raised in the prior art.

[0004] To achieve the above purpose, the present invention provides the following technical solution: An automatic control safety management method for a stacker crane, and the safety management method includes the following steps:

[0005] Step S1: Divide the shelves into multiple monitoring areas according to the storage location data information, and set a data collection period to monitor the stacker cranes in each monitoring area;

[0006] According to the floor number information of the storage location data, divide the vertical space of all the shelves in the warehouse into independent monitoring areas by floors. The monitoring area is expressed as: the set of storage locations with the same floor number in all the shelves constitutes an independent monitoring area, that is, the nth floor monitoring area is composed of the nth floor storage locations of all the shelves. The same floor storage locations of different shelves are merged into the same logical monitoring unit in the spatial dimension, and the unique identifier of each monitoring area is the floor number;

[0007] The hierarchical division method realizes the structured management of the monitoring area through logical merging in the spatial dimension, making the collection of the stacker crane status data of the same floor storage locations more targeted. The vertical hierarchical design can accurately identify the center of gravity distribution characteristics of the stacker crane at different heights. Through the horizontal comparative analysis of the vibration data of multiple stacker cranes on the same floor, the change trend of the structural stress caused by the height difference can be more intuitively found, effectively eliminating the cross-layer interference factors, and significantly improving the recognition accuracy of the equipment abnormal vibration characteristics and the timeliness of risk early warning.

[0008] Step S2: Obtain the vibration data of the stacker during the operation stay period in all monitored areas. The vibration data includes the vibration frequency and vibration amplitude of the stacker; select the stacker information with the largest vibration amplitude in each monitored area to construct a preliminary stacker analysis set.

[0009] Step S2-1: Obtain the vibration data of the stacker during the operation stay period through a vibration sensor. The operation stay period is defined as the time period when the stacker briefly stays at the current cargo position after completing the goods storage and retrieval actions. This time period includes the static moment after the stacker's load platform stops moving and the time interval of waiting for the next instruction in the static state.

[0010] Step S2-2: When no effective vibration data is generated in the monitored area during the data collection cycle, that is, when no goods storage or retrieval actions are performed by all stackers in this monitored area, mark this area as an area without operation and exclude it from the analysis; analyze the remaining monitored areas. The analysis process is as follows:

[0011] By comparing the vibration amplitudes of the stackers in each monitored area, select the stacker with the largest vibration amplitude, so that there is a unique selected stacker corresponding to each monitored area. Statistically analyze the data information of the selected stackers in each monitored area to construct a preliminary stacker analysis set.

[0012] Obtaining the stacker vibration data during the operation stay period can exclude the complex interference during the equipment movement process and ensure that the collected data can truly reflect the vibration characteristics of the stacker in the stable state. Marking and excluding the areas without operation avoids the interference of invalid data on the analysis process and improves the efficiency and accuracy of data processing. By comparing the vibration amplitudes of the stackers in each monitored area and selecting the stacker with the largest vibration amplitude to construct a preliminary analysis set, it can quickly focus on the equipment that may have potential risks.

[0013] Step S3: Set the vibration amplitude threshold according to the historical vibration data in each monitored area, and traverse the preliminary stacker analysis set to obtain the corresponding stacker vibration data in each monitored area and compare it with the vibration amplitude threshold. Screen the stackers that meet the conditions to construct an advanced vibration analysis set.

[0014] Step S3-1: Obtain the historical vibration data in each monitored area, calculate the average vibration amplitude and the standard deviation of the vibration amplitude corresponding to each area, and set the threshold coefficient in combination with the three-sigma principle to set the vibration amplitude threshold; the method of setting the vibration amplitude threshold is to obtain the vibration amplitude threshold corresponding to each monitored area according to the sum of the average vibration amplitude and the product of the standard deviation of the vibration amplitude and the threshold coefficient.

[0015] The calculation formula for setting the vibration amplitude threshold is as follows:

[0016] Tmax = T + k × b;

[0017] In the formula, T max represents the vibration amplitude threshold; T represents the average vibration amplitude corresponding to the area; b represents the standard deviation of the vibration amplitude corresponding to the area; k represents the threshold coefficient. According to the normal distribution of the three-sigma principle, the value range of the threshold coefficient k is from 1.5 to 2;

[0018] For each monitoring area, the average vibration amplitude is calculated specifically for the vibration condition of the stacker in that area. Because stackers in different areas may have different vibration amplitudes due to various factors, such as the weight of goods, the stability of the shelf structure, the workload, etc.

[0019] Step S3-2: Traverse and read the vibration amplitudes of the stackers in the preliminary stacker analysis set, compare them with the vibration amplitude thresholds set for the corresponding monitoring areas, extract the stackers whose vibration amplitudes exceed the vibration amplitude thresholds, and construct an advanced vibration analysis set;

[0020] Based on the historical vibration data of each monitoring area, the vibration amplitude threshold is set in combination with the three-sigma principle, which fully considers the statistical characteristics and fluctuation laws of the data, and can customize a reasonable judgment standard for each area. By comparing the vibration data of the stackers in the preliminary stacker analysis set with the corresponding thresholds, the stackers with excessive vibration amplitudes are screened out to construct an advanced analysis set, which further focuses on the equipment that may have potential safety hazards, improves the pertinence and accuracy of the subsequent risk assessment, avoids the ineffective analysis of a large number of normal equipment, and effectively improves the efficiency of safety management.

[0021] Step S4: Obtain the maintenance record data of each stacker to construct a maintenance analysis set, traverse and read the stacker data information in the advanced vibration analysis set, and calculate the maintenance risk value of each stacker in the advanced vibration analysis set in combination with the maintenance analysis set, and predict the vibration data in each monitoring area, and calculate the vibration risk value of each stacker according to the prediction results;

[0022] Step S4-1: The maintenance record data in the maintenance analysis set includes the maintenance timestamp and the number of maintenance times. Find the corresponding maintenance record data according to the equipment number or identification code of the stacker;

[0023] Step S4-2: Traverse the maintenance record data of each stacker in the advanced vibration analysis set, and select the time period with the largest time interval between two adjacent maintenances from the maintenance record data of multiple stackers as the maintenance time interval threshold;

[0024] Step S4-3: Subtract the timestamp of the last maintenance of the stacker from the timestamp of the current data collection cycle to obtain the operating interval after maintenance. When there is no maintenance record data for the stacker, use the timestamp of the first use of the stacker as the timestamp of the last maintenance to calculate the operating interval after maintenance;

[0025] Step S4-4: Subtract the operating interval after maintenance from the maintenance time interval threshold to obtain the predicted remaining maintenance time interval, and calculate the maintenance risk value of the stacker through the MTBF / MTTR algorithm in combination with the maintenance times of the stacker;

[0026] The maintenance risk value of the stacker is calculated using the following formula:

[0027]

[0028] In the formula, M represents the maintenance risk value of the stacker; N represents the maintenance times of the stacker; N max represents the maximum value of the maintenance times of a single stacker among all stackers; S represents the predicted remaining maintenance time interval; MTBF represents the average time between failures of the stacker; MTTR represents the average time required for the maintenance of the stacker; H avg-repair represents the average maintenance time of all stackers;

[0029] The average time between failures MTBF is calculated using the following formula:

[0030]

[0031] In the formula, t i represents the time interval from the commissioning of the i-th stacker to the current data collection cycle; n represents the number of stackers participating in the calculation of the average time between failures; N represents the maintenance times of the stacker within the time ti. The larger the value of MTBF, the higher the reliability of the stacker;

[0032] The average time required for the maintenance of the stacker MTTR is calculated using the following formula:

[0033]

[0034] In the formula, r j represents the total maintenance time spent on the j-th stacker; m represents the total maintenance time spent on all stackers. The larger the value of MTTR, the higher the possible maintenance difficulty, and the greater the calculation result included in the maintenance risk value.

[0035] Step S4-5: Obtain the corresponding historical vibration data based on the stacker equipment number or identification code in the advanced vibration analysis set, input the historical vibration data of the stacker into the time series analysis and prediction model, predict the vibration data of each stacker as the predicted vibration data, and obtain the vibration risk value according to the ratio of the predicted vibration data to the stacker vibration standard specification;

[0036] The calculation of the time series analysis and prediction model uses the following formula:

[0037]

[0038] In the formula, represents the vibration data of stacker k at the next data collection period t + 1; represents the actual vibration data of stacker k within the current data collection period t; represents the first-order difference value of stacker k, specifically the difference between two adjacent actual vibration data; represents the autoregressive coefficient of the time series analysis and prediction model; represents the moving average coefficient of the time series analysis and prediction model; represents the prediction error value of stacker k at the current data collection period t; p represents the order of the autoregressive part; q represents the moving average order;

[0039] The vibration risk value is calculated using the following formula:

[0040]

[0041] In the formula, R k represents the vibration risk value of stacker k; S max represents the upper limit value of the stacker vibration standard specification, which is the threshold that cannot be exceeded as defined by the specification;

[0042] By integrating historical data such as the maintenance timestamps and maintenance times of the stacker, and combining the MTBF / MTTR algorithm to quantitatively evaluate the equipment maintenance risk, it provides data support for maintenance decisions; the vibration prediction unit conducts trend prediction on the historical vibration data based on the time series analysis model, compares the prediction result with the standard specification to calculate the vibration risk value, and realizes the forward-looking assessment of the equipment status. The combination of the two enables the risk assessment to include both the statistical analysis of the historical maintenance rules of the equipment and the dynamic prediction of real-time vibration data, significantly improving the comprehensiveness and accuracy of safety management.

[0043] Step S5: Calculate the comprehensive risk value of each stacker crane in the advanced vibration analysis set based on the maintenance risk value and vibration risk value. Select the stacker crane with the maximum comprehensive risk value as the risk monitoring stacker crane. Extract the group of stacker cranes with risks from the vibration data analysis of the risk monitoring stacker crane as the potential risk stacker crane group. Perform vibration data prediction on the potential risk stacker crane group and monitor and give early warnings to the potential risk stacker crane group in combination with the vibration standard specifications of the stacker crane;

[0044] The comprehensive risk value of the stacker crane is calculated using the following formula:

[0045]

[0046] In the formula, represents the comprehensive risk value of stacker crane k; m1 represents the weight of the vibration risk value of stacker crane k; m2 represents the weight of the maintenance risk value of stacker crane k; and it satisfies the condition m1 + m2 = 1;

[0047] By integrating the historical maintenance records and real-time vibration data of the stacker crane, the equipment risks are evaluated step by step: First, analyze the historical data such as the maintenance time and frequency of each equipment, and calculate the maintenance risk in combination with indicators such as the mean time between failures and the average maintenance time; at the same time, use the time series model to predict the future vibration values of each equipment, and obtain the vibration risk by comparing with the safety standards; finally, comprehensively evaluate the two risks according to the weights, locate the equipment and clusters with the highest risks, and achieve dynamic early warning.

[0048] Step S5-1: Obtain the comprehensive risk value of each stacker crane through weighted fusion calculation based on the maintenance risk value and vibration risk value of each stacker crane in the analysis set;

[0049] Step S5-2: Set the feature similarity threshold, compare the comprehensive risk values of each stacker crane, and select the stacker crane with the maximum comprehensive risk value as the risk monitoring stacker crane; extract the vibration amplitude and vibration frequency data of the risk monitoring stacker crane, input the vibration data of the risk monitoring stacker crane as sample data into the neural network model for training to obtain the characteristic pattern of the vibration data, input the vibration data of other stacker cranes into the neural network model for feature analysis, and compare the feature similarity. The stacker cranes with feature similarity greater than the feature similarity threshold are used as potential risk stacker cranes, and record each potential risk stacker crane to form a potential risk stacker crane group;

[0050] The feature similarity is calculated using the following formula:

[0051]

[0052] In the formula, represents the vibration data vector of the risk monitoring stacker crane and the vibration data vector of other stacker cranes Cosine similarity; n represents the dimension of the vibration data vector, specifically the number of features included in the vibration data, including vibration amplitude and vibration frequency; xi represents the i-th component of the vibration data vector of the risk monitoring stacker ; yi represents the i-th component of the vibration data vector of other stackers ; and represents the dot product of two vectors, and the calculation method is to sum the products of the corresponding components; and respectively represent the vectors and modulus lengths of, which are calculated by taking the square root of the sum of the squares of the components.

[0053] The calculation method of setting the feature similarity threshold is the same as that of the vibration amplitude threshold, and the threshold is set by calculating the mean and standard deviation of the vibration amplitude or feature similarity during normal operation.

[0054] Step S5-3, predict the vibration data of the potential risk stacker group according to the time series analysis prediction model, and send a warning signal when the predicted vibration data exceeds the vibration standard specification of the stacker;

[0055] By integrating the maintenance records of the stacker and the real-time vibration data, first calculate the comprehensive risk value, then identify the equipment clusters with similar vibration characteristics through the neural network model, and finally conduct dynamic prediction and warning on the potential risk equipment. This method realizes the comprehensive monitoring from single equipment to cluster risks, can not only accurately locate high-risk equipment, but also discover potential hazards of similar equipment in advance through feature pattern recognition, and effectively improves the initiative of equipment maintenance and the efficiency of resource allocation by combining the prediction and warning mechanism.

[0056] Furthermore, a safety management system for the automated control of stackers, the safety management system includes: a regional division module, a vibration acquisition module, a threshold analysis module, a risk assessment module, and a warning decision module;

[0057] The regional division module is used to divide the shelf into multiple monitoring areas according to the goods location data information and set the data acquisition period; the vibration acquisition module is used to obtain the vibration data of the stacker during the operation and stay period in all monitoring areas and construct a preliminary stacker analysis set; the threshold analysis module is used to set the vibration amplitude threshold according to the historical vibration data and screen out the stackers that meet the conditions to construct an advanced vibration analysis set; the risk assessment module is used to calculate the maintenance risk value and vibration risk value of each stacker in the advanced vibration analysis set; the warning decision module is used to calculate the comprehensive risk value of the stacker, extract the potential risk stacker group and conduct monitoring and warning;

[0058] The output end of the area division module is electrically connected to the input end of the vibration acquisition module; the output end of the vibration acquisition module is electrically connected to the input end of the threshold analysis module; the output end of the threshold analysis module is electrically connected to the input end of the risk assessment module; the output end of the risk assessment module is electrically connected to the input end of the early warning decision module;

[0059] It includes a vertical stratification unit and a monitoring period unit; the vertical stratification unit is used to divide the vertical space of all the shelves in the warehouse into independent monitoring areas according to the floor number information of the storage locations; the monitoring period unit is used to set the data acquisition period for monitoring the stackers in each monitoring area;

[0060] The vibration acquisition module includes an operation stay acquisition unit and an extreme value screening unit; the operation stay acquisition unit is used to obtain the vibration data of the stacker during the operation stay period through a vibration sensor; the extreme value screening unit is used to compare the vibration amplitudes of the stackers in each monitoring area and select the stacker with the largest vibration amplitude to construct a preliminary stacker analysis set;

[0061] The threshold analysis module includes a historical data modeling unit and a threshold comparison unit; the historical data modeling unit is used to obtain the historical vibration data in each monitoring area and set the vibration amplitude threshold; the threshold comparison unit is used to compare the vibration amplitudes of the stackers in the preliminary stacker analysis set with the vibration amplitude threshold of the corresponding monitoring area and extract the stackers that exceed the threshold to construct an advanced vibration analysis set;

[0062] The risk assessment module includes a maintenance record analysis unit and a vibration prediction unit; the risk assessment module is used to calculate the maintenance risk value and the vibration risk value of each stacker in the advanced vibration analysis set; the maintenance record analysis unit is used to find the maintenance record data according to the equipment number or identification code of the stacker, calculate the operation interval after maintenance and the expected remaining maintenance time interval, and then calculate the maintenance risk value;

[0063] The early warning decision module includes a risk fusion unit and a feature clustering early warning unit; the risk fusion unit is used to calculate the comprehensive risk value through weighted fusion according to the maintenance risk value and the vibration risk value of the stacker; the feature clustering early warning unit is used to select the stacker with the largest comprehensive risk value as the risk monitoring stacker, analyze through a neural network model to find the potential risk stacker group, predict its vibration data and give an early warning when it exceeds the standard.

[0064] Compared with the prior art, the beneficial effects of the present invention are:

[0065] 1. The present invention divides the shelves into independent monitoring areas according to the layer numbers, and through the horizontal comparison of multi-dimensional data such as the vibration frequency and amplitude of multiple stackers on the same layer, effectively eliminates the cross-layer interference, and accurately identifies the change trend of the structural stress of the equipment at different heights. Combining with the three-sigma principle to dynamically set the regional vibration threshold, ensuring the efficient capture of abnormal vibration characteristics, and avoiding misjudgment or missed judgment caused by single-dimensional data.

[0066] 2. The present invention realizes the automation of equipment maintenance decision-making through the combination of historical maintenance data and real-time vibration prediction model, quantitatively analyzes the historical maintenance records based on the MTBF / MTTR algorithm, combines with the time series prediction model to dynamically evaluate the future vibration trend of the equipment, and comprehensively calculates the maintenance risk value and the vibration risk value. Obtain the comprehensive risk value through weighted fusion, provide data support for maintenance decision-making, break through the traditional lagging maintenance mode, and improve the forward-looking evaluation ability of the equipment health status.

[0067] 3. The present invention constructs a vibration characteristic pattern library, uses the cosine similarity matching technology to identify potential risk equipment clusters, and combines with the prediction and early warning mechanism to locate the hidden dangers of similar equipment in advance. This method realizes the upgrade from single-machine risk control to cluster risk monitoring, significantly improves the accuracy and response speed of maintenance resource allocation, discovers potential faults of similar equipment in advance through feature pattern recognition technology, and enhances the safety and reliability of the warehousing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 is a schematic flow chart of a method for automated control safety management of a stacker according to the present invention;

[0069] Figure 2 is a schematic structural diagram of a system for automated control safety management of a stacker according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0070] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0071] Embodiment 1: As Figure 1 shown, the present invention provides a technical solution, a method for automated control safety management of a stacker, and the safety management method includes the following steps:

[0072] Step S1: Divide the shelves into multiple monitoring areas according to the goods location data information, and set a data acquisition period to monitor the stackers in each monitoring area;

[0073] According to the layer number information of the location data, the vertical space of all the shelves in the warehouse is divided into independent monitoring areas by the number of layers. The monitoring area is represented as: the set of locations with the same layer number in all the shelves constitutes an independent monitoring area. That is, the nth layer monitoring area is composed of the nth layer locations of all the shelves. The same layer locations of different shelves are merged into the same logical monitoring unit in the spatial dimension. The unique identifier of each monitoring area is the layer number;

[0074] In specific implementation, according to the shelf height and location distribution, the vertical space is divided into several layer monitoring areas. For example, each 1 layer is a monitoring area to ensure that the same layer locations are merged logically in space. The data collection period needs to be set according to the operation frequency of the stacker. For example, it is collected once every 15 minutes, taking into account both data timeliness and system load.

[0075] Step S2: Obtain the vibration data of the stacker during the operation stay period in all the monitoring areas. The vibration data includes the vibration frequency and vibration amplitude of the stacker; Select the stacker information with the largest vibration amplitude in each monitoring area to construct a preliminary stacker analysis set;

[0076] Step S2-1: Obtain the vibration data of the stacker during the operation stay period through a vibration sensor. The operation stay period is represented as the time period when the stacker stays briefly at the current location after completing the goods storage and retrieval actions. This time period includes the stationary moment after the load platform of the stacker stops moving and the time interval of waiting for the next instruction in the stationary state;

[0077] Step S2-2: When no valid vibration data is generated in the monitoring area within the data collection period, that is, no goods picking and placing actions are performed by all the stackers in this monitoring area, mark this area as a non-operating area and exclude it from the analysis; Analyze the remaining monitoring areas. The analysis process is as follows:

[0078] Through the comparison of the vibration amplitudes of the stackers in each monitoring area, select the stacker with the largest vibration amplitude, so that there is only one selected stacker corresponding to each monitoring area. Statistically analyze the data information of the selected stackers in each monitoring area to construct a preliminary stacker analysis set;

[0079] In specific implementation, the vibration sensor should be fixed at the key stress points of the stacker frame, such as the connection between the column and the fork, to avoid interference from moving parts. The determination of the operation stay period needs to be combined with the stacker status signal. For example, the 5 seconds after the stop instruction is triggered is the stationary period. For the non-operating area, it is necessary to cross-verify through the operation logs of all the stackers in the area to prevent misjudgment. When constructing the preliminary analysis set, if the vibration amplitudes of multiple stackers are the same, preferentially select the equipment with more historical maintenance times for analysis.

[0080] Step S3: Set the vibration amplitude threshold according to the historical vibration data in each monitoring area, traverse the preliminary stacker analysis set, obtain the stacker vibration data corresponding to each monitoring area, compare it with the vibration amplitude threshold, and screen out the stackers that meet the conditions to construct an advanced vibration analysis set;

[0081] Step S3-1: Obtain the historical vibration data in each monitoring area, calculate the average vibration amplitude and the standard deviation of the vibration amplitude corresponding to each area, and set the threshold coefficient in combination with the three-sigma principle to set the vibration amplitude threshold; the method of setting the vibration amplitude threshold is to obtain the vibration amplitude threshold corresponding to each monitoring area according to the average vibration amplitude plus the product of the standard deviation of the vibration amplitude and the threshold coefficient;

[0082] Step S3-2: Traverse and read the stacker vibration amplitudes in the preliminary stacker analysis set, compare them with the vibration amplitude thresholds set corresponding to the monitoring areas, extract the stackers whose vibration amplitudes exceed the vibration amplitude thresholds, and construct an advanced vibration analysis set;

[0083] In specific implementation, the historical vibration data needs to cover multiple complete operation cycles, and calculate the average vibration amplitude and standard deviation of each monitoring area. The value of the threshold coefficient k needs to be dynamically adjusted according to the equipment type. For example, for heavy-duty stackers, k = 1.8, and for light-duty ones, k = 1.5. When traversing and screening, accidental impacts need to be excluded, such as the instantaneous vibration generated by cargo collision. Such vibration data will be more prominent and will not be used as the analysis data source. Abnormalities are reconfirmed through continuous multiple above-standard data.

[0084] Step S4: Obtain the maintenance record data of each stacker to construct a maintenance analysis set, traverse and read the stacker data information in the advanced vibration analysis set, calculate the maintenance risk value of each stacker in the advanced vibration analysis set in combination with the maintenance analysis set, and predict based on the vibration data in each monitoring area, and calculate the vibration risk value of each stacker according to the prediction result;

[0085] Step S4-1: The maintenance record data in the maintenance analysis set includes the maintenance timestamp and the number of maintenance times. Find the corresponding maintenance record data according to the equipment number or identification code of the stacker;

[0086] Step S4-2: Traverse the maintenance record data of each stacker in the advanced vibration analysis set, and select the time period with the largest time interval between two adjacent maintenance times from the maintenance record data of multiple stackers as the maintenance time interval threshold;

[0087] Step S4-3: Subtract the timestamp of the stacker's most recent maintenance from the timestamp of the current data collection cycle to obtain the operating interval after maintenance. When there is no maintenance record data for the stacker, use the timestamp of the stacker's first commissioning as the timestamp of the most recent maintenance to calculate the operating interval after maintenance;

[0088] Step S4-4: Subtract the operating interval after maintenance from the maintenance interval threshold to obtain the predicted remaining maintenance time interval, and calculate the maintenance risk value of the stacker through the MTBF / MTTR algorithm in combination with the number of maintenance times of the stacker;

[0089] Step S4-5: Obtain the corresponding historical vibration data according to the stacker equipment number or identification code in the advanced vibration analysis set, input the historical vibration data of the stacker into the time series analysis prediction model, predict the vibration data of each stacker as the predicted vibration data, and obtain the vibration risk value according to the ratio of the predicted vibration data to the stacker vibration standard specification;

[0090] In specific implementation, the maintenance record data needs to include detailed information such as fault codes and maintenance contents, and the MTBF / MTTR algorithm needs to be corrected in combination with industry standard parameters. The time series analysis prediction model, such as ARIMA(2,1,1), regularly updates historical data.

[0091] Step S5: Calculate the comprehensive risk value of each stacker in the advanced vibration analysis set according to the maintenance risk value and the vibration risk value, select the stacker with the largest comprehensive risk value as the risk monitoring stacker, and extract the stacker group with risks from the vibration data analysis of the risk monitoring stacker as the potential risk stacker group. Predict the vibration data of the potential risk stacker group, and monitor and give early warnings to the potential risk stacker group in combination with the stacker vibration standard specification;

[0092] Step S5-1: Calculate the comprehensive risk value of each stacker through weighted fusion according to the maintenance risk value and the vibration risk value of each stacker in the analysis set;

[0093] Step S5-2: Set the feature similarity threshold, compare the comprehensive risk values of each stacker, and select the stacker with the largest comprehensive risk value as the risk monitoring stacker; Extract the vibration amplitude and vibration frequency data of the risk monitoring stacker, input the vibration data of the risk monitoring stacker into the neural network model for training to obtain the feature pattern of the vibration data, input the vibration data of other stackers into the neural network model for feature analysis, and compare the feature similarity. The stacker with a feature similarity greater than the feature similarity threshold is used as the potential risk stacker, and record each potential risk stacker to form a potential risk stacker group;

[0094] Step S5-3: Predict the vibration data of the potential risk stacker group according to the time series analysis prediction model. When the predicted vibration data exceeds the vibration standard specification of the stacker, an early warning signal is issued;

[0095] In specific implementation, the neural network model uses a convolutional neural network (CNN) to extract the frequency domain features of the vibration signal. The training samples need to include historical data of normal and faulty states. The feature similarity threshold is determined by cross-validation. For example, when the cosine similarity > 0.85, the early warning information needs to be synchronized to the operation and maintenance terminal and the processing result is recorded.

[0096] Embodiment 2, as Figure 2 shown, the present invention provides a safety management system for the automatic control of stackers, and the safety management system includes: a region division module, a vibration acquisition module, a threshold analysis module, a risk assessment module, and an early warning decision module;

[0097] The region division module is used to divide the shelves into multiple monitoring regions according to the goods location data information and set the data acquisition period; the vibration acquisition module is used to obtain the vibration data of the stackers during the operation and stay period in all monitoring regions and construct a preliminary stacker analysis set; the threshold analysis module is used to set the vibration amplitude threshold according to the historical vibration data and screen out the stackers that meet the conditions to construct an advanced vibration analysis set; the risk assessment module is used to calculate the maintenance risk value and vibration risk value of each stacker in the advanced vibration analysis set; the early warning decision module is used to calculate the comprehensive risk value of the stackers, extract the potential risk stacker group and conduct monitoring and early warning;

[0098] The output end of the region division module is electrically connected to the input end of the vibration acquisition module; the output end of the vibration acquisition module is electrically connected to the input end of the threshold analysis module; the output end of the threshold analysis module is electrically connected to the input end of the risk assessment module; the output end of the risk assessment module is electrically connected to the input end of the early warning decision module;

[0099] The region division module includes a vertical stratification unit and a monitoring period unit; the vertical stratification unit is used to divide the vertical space of all the shelves in the warehouse into independent monitoring regions according to the layer number information of the goods location; the monitoring period unit is used to set the data acquisition period for monitoring the stackers in each monitoring region;

[0100] The vibration acquisition module includes an operation and stay acquisition unit and an extreme value screening unit; the operation and stay acquisition unit is used to obtain the vibration data of the stacker during the operation and stay period through a vibration sensor; the extreme value screening unit is used to compare the vibration amplitudes of the stackers in each monitoring region and select the stacker with the largest vibration amplitude to construct a preliminary stacker analysis set;

[0101] The threshold analysis module includes a historical data modeling unit and a threshold comparison unit; the historical data modeling unit is used to obtain the historical vibration data in each monitoring area and set the vibration amplitude threshold; the threshold comparison unit is used to compare the vibration amplitude of the stacker in the preliminary stacker analysis set with the vibration amplitude threshold of the corresponding monitoring area, and extract the stackers that exceed the threshold to construct an advanced vibration analysis set;

[0102] The risk assessment module includes a maintenance record analysis unit and a vibration prediction unit; the risk assessment module is used to calculate the maintenance risk value and vibration risk value of each stacker in the advanced vibration analysis set; the maintenance record analysis unit is used to find the maintenance record data according to the equipment number or identification code of the stacker, calculate the running interval after maintenance and the predicted remaining maintenance time interval, and then calculate the maintenance risk value;

[0103] The warning decision module includes a risk fusion unit and a feature clustering warning unit; the risk fusion unit is used to calculate the comprehensive risk value through weighted fusion according to the maintenance risk value and vibration risk value of the stacker; the feature clustering warning unit is used to select the stacker with the largest comprehensive risk value as the risk monitoring stacker, analyze and find the potential risk stacker group through the neural network model, predict its vibration data and give a warning when it exceeds the standard.

[0104] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claimed claims.

Claims

1. A method for automatic control safety management of a stacker, characterized in that: The safety management method comprises the following steps: Step S1, dividing the shelf into multiple monitoring areas according to the cargo location data information, and setting a data collection cycle to monitor the stacker in each monitoring area; Step S2, obtaining vibration data of the stackers in all monitoring areas during the operation dwell period, wherein the vibration data includes the vibration frequency and vibration amplitude of the stackers; selecting the stacker information with the largest vibration amplitude in each monitoring area to construct a preliminary stacker analysis set; Step S3, setting a vibration amplitude threshold according to the historical vibration data in each monitoring area, and traversing the preliminary stacker analysis set, obtaining the corresponding stacker vibration data in each monitoring area and comparing it with the vibration amplitude threshold, and screening the stackers that meet the conditions to construct an advanced vibration analysis set; Step S4, obtaining the maintenance record data of each stacker to construct a maintenance analysis set, traversing and reading the advanced vibration analysis set to obtain the stacker data information, and combining the maintenance analysis set to calculate the maintenance risk value of each stacker in the advanced vibration analysis set, and predicting the vibration data in each monitoring area, and calculating the vibration risk value of each stacker according to the prediction results; Step S5: Calculate the comprehensive risk value of each stacker in the advanced vibration analysis set based on the maintenance risk value and the vibration risk value, select the stacker with the largest comprehensive risk value as the risk monitoring stacker, and extract the stacker group with risks based on the vibration data analysis of the risk monitoring stacker as the potential risk stacker group, predict the vibration data of the potential risk stacker group, and monitor and warn the potential risk stacker group in combination with the stacker vibration standard specification.

2. A method for automatic control safety management of stacker crane according to claim 1, characterized in that: In step S1, according to the layer number information of the cargo location data, the vertical space of all shelves in the warehouse is divided into independent monitoring areas according to the number of layers. The monitoring area is expressed as follows: the set of cargo locations with the same layer number in all shelves constitutes an independent monitoring area, that is, the n-th layer monitoring area is composed of the n-th layer cargo locations of all shelves, and the same-layer cargo locations of different shelves are merged into the same logical monitoring unit in the spatial dimension. The unique identifier of each monitoring area is the layer number.

3. A method for automatic control safety management of stacker crane according to claim 2, characterized in that: The specific steps of step S2 are as follows: Step S2-1, obtaining vibration data of the stacker during the operation dwell period through a vibration sensor, wherein the operation dwell period is the time period during which the stacker briefly stays at the current cargo location after completing the cargo storage and retrieval action, and this time period includes the static moment after the cargo loading platform of the stacker stops moving and the time interval during which the stacker waits for the next instruction in the static state; Step S2-2: When no valid vibration data is generated in the monitoring area during the data collection period, that is, all stackers have no cargo picking and placing actions in the monitoring area, the area is marked as a non-operating area and excluded from the analysis; the remaining monitoring areas are analyzed, and the analysis process is as follows: By comparing the vibration amplitudes of the stackers in each monitoring area, the stacker with the largest vibration amplitude is selected so that there is only one selected stacker in any monitoring area. The data information of the stackers selected in each monitoring area is statistically analyzed to construct a preliminary stacker analysis set.

4. The method for automatic control safety management of a stacker according to claim 3, characterized in that: The specific steps of step S3 are as follows: Step S3-1, obtaining historical vibration data in each monitoring area, and calculating the vibration amplitude average value and vibration amplitude standard deviation corresponding to each area, and setting the vibration amplitude threshold by setting the threshold coefficient in combination with the three sigma principle; The method for setting the vibration amplitude threshold is to obtain the vibration amplitude threshold corresponding to each monitoring area according to the average vibration amplitude plus the product of the vibration amplitude standard deviation and the threshold coefficient; Step S3-2, traverse and read the stacker vibration amplitudes in the preliminary stacker analysis set, compare with the vibration amplitude threshold set corresponding to the monitoring area, extract the stackers whose vibration amplitude exceeds the vibration amplitude threshold, and construct an advanced vibration analysis set.

5. The method for automatic control safety management of stacker crane according to claim 4, characterized in that: The specific steps of step S4 are as follows: Step S4-1, the maintenance record data in the maintenance analysis set includes a maintenance timestamp and a maintenance number, and the corresponding maintenance record data is searched according to the equipment number or identity identification code of the stacker; Step S4-2, traversing the maintenance record data of each stacker in the advanced vibration analysis set, and selecting the time period with the largest time interval between two adjacent maintenances from the maintenance record data of multiple stackers as the maintenance time interval threshold; Step S4-3, use the timestamp data of the current data collection cycle minus the timestamp of the stacker's most recent maintenance to obtain the post-maintenance operation interval. When the stacker has no maintenance record data, use the timestamp of the stacker's first use as the most recent maintenance timestamp to calculate the post-maintenance operation interval.

6. A method for automatic control safety management of stacker crane according to claim 5, characterized in that: In step S4, it also includes: Step S4-4, using the maintenance time interval threshold minus the post-maintenance operation interval to obtain the estimated remaining maintenance time interval, and combining the number of maintenance times of the stacker crane with the MTBF / MTTR algorithm to calculate the maintenance risk value of the stacker crane; Step S4-5, obtain the corresponding historical vibration data according to the stacker equipment number or identity identification code in the advanced vibration analysis set, input the historical vibration data of the stacker into the time series analysis prediction model, predict the vibration data of each stacker as the predicted vibration data, and obtain the vibration risk value according to the ratio of the predicted vibration data to the stacker vibration standard specification.

7. A method for automatic control safety management of stacker according to claim 6, characterized in that: The specific steps of step S5 are as follows: Step S5-1, obtaining a comprehensive risk value of each stacker by weighted fusion calculation based on the maintenance risk value and vibration risk value of each stacker in the analysis set; Step S5-2, setting a feature similarity threshold, comparing the comprehensive risk values ​​of each stacker, and selecting the stacker with the largest comprehensive risk value as the risk monitoring stacker; Extract the vibration amplitude and vibration frequency data of the risk monitoring stacker, input the vibration data of the risk monitoring stacker as sample data into the neural network model for training to obtain the characteristic pattern of the vibration data, input the vibration data of other stackers into the neural network model for feature analysis, and by comparing the feature similarity, take the stackers with feature similarity greater than the feature similarity threshold as potential risk stackers, store and record each potential risk stacker to form a potential risk stacker group; Step S5-3: predict vibration data of the potential risk stacker group according to the time series analysis prediction model, and issue a warning signal when the predicted vibration data exceeds the stacker vibration standard specification.

8. A safety management system for automatic control of a stacker, which is applied to a safety management method for automatic control of a stacker according to any one of claims 1 to 7, characterized in that: The safety management system includes: a region division module, a vibration collection module, a threshold analysis module, a risk assessment module and an early warning decision module; The area division module is used to divide the shelf into multiple monitoring areas according to the cargo location data information, and set the data collection cycle; the vibration collection module is used to obtain the vibration data of the stacker in all monitoring areas during the operation stay period and build a preliminary stacker analysis set; the threshold analysis module is used to set the vibration amplitude threshold according to the historical vibration data, and filter out the stacker that meets the conditions to build an advanced vibration analysis set; the risk assessment module is used to calculate the maintenance risk value and vibration risk value of each stacker in the advanced vibration analysis set; the early warning decision module is used to calculate the comprehensive risk value of the stacker, extract the potential risk stacker group and perform monitoring and early warning; The output end of the area division module is electrically connected to the input end of the vibration collection module; the output end of the vibration collection module is electrically connected to the input end of the threshold analysis module; the output end of the threshold analysis module is electrically connected to the input end of the risk assessment module; the output end of the risk assessment module is electrically connected to the input end of the early warning decision module.

9. The automatic control safety management system for stacking crane according to claim 8, characterized in that: The system comprises a vertical stratification unit and a monitoring cycle unit; the vertical stratification unit is used to divide the vertical space of all shelves in the warehouse into independent monitoring areas according to the number of layers according to the layer number information of the cargo location; the monitoring cycle unit is used to set the data collection cycle for monitoring the stacker in each monitoring area; The vibration collection module includes an operation stop collection unit and an extreme value screening unit; the operation stop collection unit is used to obtain the vibration data of the stacker during the operation stop period through a vibration sensor; the extreme value screening unit is used to compare the vibration amplitude of the stacker in each monitoring area, and select the stacker with the largest vibration amplitude to construct a preliminary stacker analysis set; The threshold analysis module includes a historical data modeling unit and a threshold comparison unit; the historical data modeling unit is used to obtain historical vibration data in each monitoring area and set a vibration amplitude threshold; the threshold comparison unit is used to compare the vibration amplitude of the stacker in the preliminary stacker analysis set with the vibration amplitude threshold of the corresponding monitoring area, and extract the stackers exceeding the threshold to construct an advanced vibration analysis set.

10. The automatic control safety management system for stacking crane according to claim 8, characterized in that: The risk assessment module includes a maintenance record analysis unit and a vibration prediction unit; the risk assessment module is used to calculate the maintenance risk value and vibration risk value of each stacker in the advanced vibration analysis set; The maintenance record analysis unit is used to search for maintenance record data according to the equipment number or identity code of the stacker, calculate the operation interval after maintenance and the estimated remaining maintenance time interval, and then calculate the maintenance risk value; The early warning decision module includes a risk fusion unit and a feature clustering early warning unit; The risk fusion unit is used to obtain a comprehensive risk value through weighted fusion calculation according to the maintenance risk value and the vibration risk value of the stacker; The feature clustering warning unit is used to select the stacker with the largest comprehensive risk value as the risk monitoring stacker, find out the potential risk stacker group through neural network model analysis, predict its vibration data and issue a warning when it exceeds the standard.

Images