Intelligent matching system for dynamic speed regulation of conveyor belt for flexible production

Abstract

The invention discloses an intelligent matching system for dynamic speed regulation of a conveyor belt for flexible production, and relates to the technical field of speed regulation of conveyor belts, and the matching system comprises a conveying data analysis module which is used for carrying out the data collection of each conveying process of the conveyor belt, and generating a conveying log; identifying congestion behaviors existing in any transmission log to obtain a congestion behavior set; the congestion characteristic analysis module is used for carrying out difference comparison on each congestion behavior and carrying out category division; carrying out abnormal feature identification on the congestion behavior; the transmission prediction decision module is used for constructing a congestion prediction model to predict the congestion condition of the transmission log; making corresponding adjustment decisions for various congestion behaviors; the real-time prediction feedback module is used for monitoring the real-time conveying process on the conveying belt to obtain a real-time conveying log, and performing congestion prediction and timely adjustment on the real-time conveying log; and abnormal conditions existing after adjustment are identified and reminded.

Description

Technical Field

[0001] The present invention relates to the technical field of conveyor belt speed regulation, and specifically to an intelligent matching system for dynamic speed regulation of a conveyor belt for flexible production. Background Art

[0002] A conveyor belt for flexible production is an intelligent material transmission device used in modern intelligent manufacturing systems. Its core design goal is to adapt to the production requirements of multi-variety, small-batch, and rapid production changeover. By dynamically adjusting speed, path, and operation mode, it maximizes the flexibility and efficiency of the production process.

[0003] Traditional conveyor belt systems mostly adopt fixed speed regulation or speed regulation strategies based on simple rules, such as preset speed curves. They cannot adapt to dynamic changes in the production environment in real time. The system is difficult to quickly adjust the conveyor speed, with serious feedback delays, resulting in untimely speed regulation responses and affecting production efficiency. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent matching system for dynamic speed regulation of a conveyor belt for flexible production to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An intelligent matching system for dynamic speed regulation of a conveyor belt for flexible production, the matching system includes a conveyor data analysis module, a congestion feature analysis module, a conveyor prediction decision module, and a real-time prediction feedback module.

[0006] The conveyor data analysis module is used to collect data on each conveyor process of the conveyor belt through preset monitoring devices to generate a conveyor log; identify congestion behaviors existing in any conveyor log to obtain a congestion behavior set of any conveyor log.

[0007] The congestion feature analysis module is used to compare the differences between each congestion behavior in all congestion behavior sets, classify different congestion behaviors; identify abnormal features of any type of congestion behavior and analyze the feature ranges of each abnormal feature.

[0008] The conveyor prediction decision module is used to construct a congestion prediction model to predict the congestion situation of the conveyor log and train based on the actual congestion situation of any conveyor log; formulate corresponding adjustment decisions for various congestion behaviors based on the prediction results.

[0009] The real-time prediction feedback module is used to monitor the real-time conveyor process on the conveyor belt to obtain a real-time conveyor log, predict congestion for the real-time conveyor log and make timely adjustments; monitor the change situation of abnormal features after adjustment, identify existing abnormal situations and give reminders.

[0010] Further, the conveyor data analysis module includes a conveyor log collection unit and a conveyor congestion identification unit;

[0011] The conveyor log collection unit is used to divide the conveyor belt into several conveyor areas according to the preset function types, and corresponding monitoring devices are installed for each conveyor area; whenever the conveyor belt starts to convey, each monitoring device starts to collect conveyor data for the corresponding conveyor area until the conveyor belt stops conveying; the monitoring data collected during each complete conveying process is presented according to the monitoring time to show the data change situation, and the corresponding conveyor log is obtained; the conveyor belt can be divided into multiple conveyor areas such as a handling section, a sorting section, a buffer section, etc. according to the function types, and pressure sensors, speed sensors, monitors and other monitoring devices are installed for each conveyor area to collect data during the conveying process;

[0012] The conveyor congestion identification unit is used to arbitrarily select a conveyor log, then arbitrarily select a monitoring device from the selected conveyor log, extract the monitoring data of the monitoring device at any monitoring time point, set the monitoring data monitored by each monitoring device as a type of monitoring data, and preset a normal value range for each type of monitoring data. Among them, a normal value range r=(d1, d2) is preset for the selected monitoring device, where d1 is the minimum value and d2 is the maximum value, and the value of the monitoring data at the monitoring time point t0 is set as d t0 , if then obtain several adjacent and continuous monitoring time points adjacent to the monitoring time point t0. If the values of the monitoring data at the several monitoring time points do not belong to the normal value range, then merge the several monitoring time points to obtain a time interval, and obtain the conveying behavior within the time interval to get a congestion behavior of the conveyor log; obtain the congestion behaviors of each time interval to get the congestion behavior set of the selected conveyor log; during the conveying process of the conveyor belt, data volatility will occur. Therefore, the monitoring data at only one time point cannot accurately reflect whether there is a congestion behavior. Only when there is a continuous time interval with anomalies can it be determined as a congestion behavior.

[0013] Further, the congestion feature analysis module includes a conveyor congestion division unit and an abnormal feature identification unit;

[0014] The conveyor congestion division unit is used to arbitrarily select a congestion behavior from a certain conveyor log, and obtain the time interval and various types of monitoring data where the selected congestion behavior is located; compare the differences in the time intervals and monitoring data between any two congestion behaviors to determine whether to classify the two user behaviors;

[0015] An abnormal feature recognition unit is used to analyze the abnormal conditions of any congestion behavior in various types of monitoring data for any same type, and extract a number of abnormal features; and extract the monitoring data corresponding to any abnormal feature in different congestion behaviors to obtain the feature value range of any abnormal feature.

[0016] Further, the conveying congestion division unit includes:

[0017] Arbitrarily select a certain conveying log, and arbitrarily select a congestion behavior from the certain conveying log, set the selected congestion behavior as the target congestion behavior, and obtain the time interval of the target congestion behavior as (t g1 , t g2 );

[0018] Obtain the conveying rates of the conveyor belt in each conveying area to obtain the expected time intervals of each conveying area. Among them, set the expected time interval of the i-th conveying area as (t1 i , t2 i ); if (t g1 , t g2 ) ∈ (t1 i , t2 i ), then obtain that the expected occurrence area of the target congestion behavior is the i-th conveying area, and use the monitoring device to capture the actual occurrence area of the target congestion behavior as the a-th conveying area. Set a conveying deviation mark F. If i ≠ a, then set the first deviation mark F (i,a) such that F = F (i,a) , if i = a, then set the second deviation mark F i such that F = F i ; The first deviation mark indicates that due to the congestion behavior, the conveyed products are congested in other conveying areas, indicating that the conveying process is blocked; the second deviation mark indicates the congestion situation in a conveying area; since the functional types of different conveying areas are different, the congestion behaviors when congested in different areas are different. Therefore, it is possible to judge whether the congestion behaviors are of the same type by the difference in the congestion areas;

[0019] Arbitrarily select the presentation of the monitoring data of a monitoring device changing with the monitoring time, obtain the preset normal value range of the selected monitoring device. If there is a continuous time interval in which the monitoring values of any monitoring time point do not belong to the normal value range, then perform an abnormal mark on the selected monitoring device to obtain the set A of monitoring devices with abnormal marks included in the target congestion behavior;

[0020] Select a congestion behavior from the remaining congestion behaviors and set it as the comparison congestion behavior, and obtain the conveying deviation mark F ’ of the comparison congestion behavior and the set A of all monitoring devices with abnormal marks’ , if F ’ = F and A = A ’ , then classify the target congestion behavior and the comparison congestion behavior as the same type of congestion behavior; otherwise, classify them as different congestion behaviors.

[0021] Furthermore, the abnormal feature recognition unit includes:

[0022] Arbitrarily select a group of the same type of congestion behaviors, and extract the set of monitoring devices with abnormal marks in the same type of congestion behaviors; arbitrarily select a monitoring device from the set of monitoring devices, and obtain the monitoring data presented by the monitoring device in all transportation logs to obtain a complete numerical range (d min , d max ). Set the normal numerical range of the monitoring device as (d1, d2). If d1 > d min , then obtain a deviation numerical range p1 = (d min , d1). If d2 < d max , then obtain a deviation numerical range p2 = (d2, d max );

[0023] If there is a deviation numerical range for the selected monitoring device, then set the data type of the monitoring data as the abnormal feature of the selected monitoring device; extract the abnormal features of each monitoring device in each transportation log containing the same type of congestion behaviors to obtain several abnormal features of the same type of congestion behaviors;

[0024] Arbitrarily select an abnormal feature, obtain the transportation log where any congestion behavior in the same type of congestion behaviors is located, and obtain the deviation numerical range (d1 ’ , d2 ’ ) of the monitoring device where the selected abnormal feature is located. If d1 ’ < d2 < d2 ’ , then obtain the actual deviation range of the transportation log where it is located as (d2, d2 ’ ); extract and take the intersection of the actual deviation ranges of each congestion behavior in the same type of congestion behaviors to obtain the feature numerical range of the selected abnormal feature; there will be various congestion behaviors in the same transportation area, caused by various reasons. Therefore, the deviation situations of the monitoring data caused by different congestion behaviors are different. In order to distinguish different congestion behaviors, it is necessary to further subdivide the deviation situations to better provide reference data for the subsequent prediction model.

[0025] Furthermore, the transportation prediction decision module includes a prediction model construction unit and an adjustment decision making unit;

[0026] A prediction model construction unit is used to arbitrarily select a transportation log, analyze the deviation situation of abnormal features included in any transportation area in the selected transportation log, analyze the abnormal occurrence frequency of each abnormal feature, obtain the abnormal weight of each abnormal feature, construct a congestion prediction model to predict the congestion risk of each transportation area; obtain a risk prediction threshold for judging whether there is a congestion behavior in the transportation area through the actual congestion situation of each transportation log in each transportation area.

[0027] An adjustment decision-making unit is used to extract abnormal features from the transportation areas in any transportation log where the congestion risk value exceeds the risk prediction threshold, adjust the numerical values of the monitoring data corresponding to any abnormal feature, and generate an adjustment decision.

[0028] Furthermore, the prediction model construction unit includes:

[0029] Arbitrarily select a transportation area from the selected transportation log, obtain several abnormal features included in the transportation area, and obtain the feature value range of each abnormal feature. Among them, it is set that the feature value range of the j-th abnormal feature in the transportation area is [d j (min), d j (max)]; obtain the normal value range (d1 j , d2 j ) of the j-th abnormal feature, and set a feature deviation value Q j for the j-th abnormal feature. If d j (max) < d1 j , then Q j = d j (max). If d2 j < d j (min), then Q j = d j (min); obtain the deviation amplitude η j = [Min(|Q j - d1 j |, |Q j - d2 j |)] / [Select(d1 j , d2 j )], where Min() is the minimum value function and Select() is the selection function. If the minimum value function obtains |Q j - d1 j |, then Select(d1 j , d2 j ) = d1 j . If the minimum value function obtains |Q j - d2 j|, then Select(d1 j , d2 j ) = d2 j ; By selecting the value with the smallest deviation degree as the basic situation for risk prediction, because if there is no adjustment in the case of the smallest deviation degree, congestion behavior will occur;

[0030] Obtain the occurrence times of the j-th abnormal feature in all transportation logs as m j and the number of behaviors containing the j-th abnormal feature in all congestion behaviors as n j , calculate the abnormal weight f of the j-th abnormal feature j = n j / m j ;

[0031] Construct a congestion prediction model to calculate the congestion risk prediction value R of the transportation area ex :

[0032]

[0033] If there is congestion behavior in the selected transportation area in the selected transportation logs, then set the congestion risk prediction value R ex as the abnormal prediction value; Obtain the abnormal prediction values of any transportation area where congestion behavior is located, and select the smallest abnormal prediction value as the risk prediction threshold for judging whether there is congestion behavior in the transportation area.

[0034] Furthermore, adjust the decision-making unit, including:

[0035] Set the risk prediction threshold as R th , select a transportation area from any transportation log, if the congestion risk prediction value obtained in the selected transportation area is R ex , if R ex ≥R th , then extract all abnormal features included in the selected transportation area;

[0036] Obtain the monitoring device where any abnormal feature is located, formulate an adjustment plan to adjust the conveyor belt in the selected transportation area, so that the monitoring data range of the monitoring device is within the normal value range; formulate corresponding adjustment plans for all abnormal features to generate an adjustment decision for the selected transportation area.

[0037] Furthermore, the real-time prediction feedback module includes a real-time transportation prediction unit and an abnormal congestion feedback unit;

[0038] A real-time output prediction unit, which is used to obtain the monitoring data of each monitoring device in the real-time conveyor log, so as to obtain the monitoring data of any monitoring device at the current monitoring time point. If the monitoring data at the current monitoring time point does not belong to the preset normal value range, abnormal feature extraction is performed from the monitoring device, and the deviation amplitude η of the monitoring data is obtained. now And the abnormal weight f of the extracted abnormal features; obtain the abnormal weights of each abnormal feature and the deviation amplitude of the corresponding monitoring data in any conveyor area, and input them into the congestion prediction model to obtain the congestion risk prediction value of each conveyor area in the real-time conveyor log, and adjust the conveyor areas whose congestion risk prediction values exceed the risk prediction threshold.

[0039] An abnormal congestion feedback unit, which is used to obtain the monitoring data of each monitoring device in the adjusted conveyor area every unit time. If the monitoring data of one monitoring device does not belong to the normal value range in a continuous number of monitoring time points, an abnormal reminder is sent to the adjusted conveyor area.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] 1. By analyzing the congestion behavior existing in the conveyor belt during the conveying process, the present invention constructs a prediction model to predict the congestion of the conveying state at any moment, helps to timely adjust the conveying speed of the conveyor belt in each link, avoids the situation of untimely speed adjustment response, and effectively improves production efficiency.

[0042] 2. By classifying and analyzing the congestion behaviors existing in the historical conveyor log and setting the identification methods for different types of congestion behaviors, the present invention can timely capture potential congestion situations during the conveying process and give timely feedback to help the staff make timely corrections.

[0043] 3. By formulating corresponding adjustment strategies for the conveyor belt, the present invention can well adapt to the dynamic changes brought by the production environment, enable the conveyor belt to quickly perform feedback adjustment when detecting congestion abnormalities at any moment, and avoid the interruption of the production process. Description of the Drawings

[0044] Figure 1 It is a schematic structural diagram of a flexible production conveyor belt dynamic speed regulation intelligent matching system. Detailed Embodiments

[0045] Next, in combination with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0046] Embodiment: As Figure 1 shown, the present invention provides a dynamic speed regulation intelligent matching system for a conveyor belt used in flexible production. The matching system includes a conveyor belt data analysis module, a congestion feature analysis module, a conveyor belt prediction and decision-making module, and a real-time prediction and feedback module;

[0047] The conveyor belt data analysis module is used to collect data on each conveying process of the conveyor belt through a preset monitoring device to generate a conveying log; identify congestion behaviors existing in any conveying log to obtain a congestion behavior set of any conveying log;

[0048] Among them, the conveyor belt data analysis module includes a conveying log collection unit and a conveying congestion identification unit;

[0049] The conveying log collection unit is used to divide the conveyor belt into several conveying areas according to a preset function type, and install corresponding monitoring devices for each conveying area; whenever the conveyor belt starts to convey, each monitoring device starts to collect conveying data for the corresponding conveying area until the conveyor belt stops conveying; present the change of the monitoring data collected during each complete conveying process according to the monitoring time to obtain the corresponding conveying log;

[0050] The conveying congestion identification unit is used to arbitrarily select a conveying log, arbitrarily select a monitoring device from the selected conveying log, extract the monitoring data of the monitoring device at any monitoring time point, set the monitoring data monitored by each monitoring device as a type of monitoring data, and preset a normal value range for each type of monitoring data. Among them, a normal value range r=(d1, d2) is preset for the selected monitoring device, where d1 is the minimum value and d2 is the maximum value. Set the value of the monitoring data at the monitoring time point t0 as d t0 If then obtain several adjacent and continuous monitoring time points adjacent to the monitoring time point t0. If the values of the monitoring data at the several monitoring time points do not belong to the normal value range, merge the several monitoring time points to obtain a time interval, and obtain the conveying behavior in the time interval to obtain a congestion behavior of the conveying log; obtain the congestion behaviors of each time interval to obtain the congestion behavior set of the selected conveying log.

[0051] The congestion feature analysis module is used to compare the differences between various congestion behaviors in the set of all congestion behaviors, classify different congestion behaviors into categories; identify abnormal features for any type of congestion behavior, and analyze the feature ranges of each abnormal feature;

[0052] Among them, the congestion feature analysis module includes a transportation congestion classification unit and an abnormal feature identification unit;

[0053] The transportation congestion classification unit is used to arbitrarily select a congestion behavior from a certain transportation log, obtain the time interval and various monitoring data where the selected congestion behavior is located; compare the differences in the time intervals and monitoring data between any two congestion behaviors to determine whether to classify the two user behaviors into categories;

[0054] The abnormal feature identification unit is used to analyze the abnormal conditions of any congestion behavior in various monitoring data for any same category, and extract several abnormal features; and extract the monitoring data corresponding to any abnormal feature in different congestion behaviors to obtain the feature value range of any abnormal feature.

[0055] Among them, the transportation congestion classification unit includes:

[0056] Arbitrarily select a certain transportation log, arbitrarily select a congestion behavior from the certain transportation log, set the selected congestion behavior as the target congestion behavior, and obtain the time interval of the target congestion behavior as (t g1 ,t g2 );

[0057] Obtain the transportation rates of the conveyor belt in each transportation area to obtain the expected time intervals of each transportation area. Among them, set the expected time interval of the i-th transportation area as (t1 i ,t2 i ); if (t g1 ,t g2 ) ∈ (t1 i ,t2 i ), then obtain that the expected occurrence area of the target congestion behavior is the i-th transportation area, use the monitoring device to capture the actual occurrence area of the target congestion behavior as the a-th transportation area, set a transportation deviation mark F, if i ≠ a, then set the first deviation mark F (i,a) such that F = F (i,a) , if i = a, then set the second deviation mark F i such that F = F i ;

[0058] Arbitrarily select the presentation of the monitoring data of a monitoring device over time, obtain the preset normal value range of the selected monitoring device. If there is a continuous time interval in which the monitoring values at any monitoring time point do not belong to the normal value range, then mark the selected monitoring device as abnormal, and obtain the set A of monitoring devices with abnormal marks included in the target congestion behavior;

[0059] Re-select a congestion behavior from the remaining congestion behaviors and set it as the comparison congestion behavior, and obtain the delivery deviation mark F of the comparison congestion behavior ’ and the set A of all monitoring devices with abnormal marks ’ , if F ’ =F and A = A ’ , then classify the target congestion behavior and the comparison congestion behavior as the same type of congestion behavior, otherwise, classify them as different congestion behaviors.

[0060] Among them, the abnormal feature recognition unit includes:

[0061] Arbitrarily select a group of the same type of congestion behaviors, and extract the set of monitoring devices with abnormal marks in the same type of congestion behaviors; arbitrarily select a monitoring device from the set of monitoring devices, and obtain the monitoring data presented by the monitoring device in all delivery logs, and obtain a complete value range (d min , d max ). Set the normal value range of the monitoring device as (d1, d2). If d1 > d min , then obtain a deviation value range p1=(d min , d1), if d2 < d max , then obtain a deviation value range p2=(d2, d max );

[0062] If there is a deviation value range for the selected monitoring device, then set the data type of the monitoring data as the abnormal feature of the selected monitoring device; extract the abnormal features of each monitoring device in each delivery log containing the same type of congestion behaviors, and obtain several abnormal features of the same type of congestion behaviors;

[0063] Arbitrarily select an abnormal feature, obtain the delivery log where any congestion behavior in the same type of congestion behaviors is located, and obtain the deviation value range (d1 ’ , d2 ’ ) of the monitoring device where the selected abnormal feature is located. If d1 ’ < d2 < d2 ’ , then obtain the actual deviation range of the delivery log where it is located as (d2, d2 ’ ); extract and take the intersection of the actual deviation ranges of each congestion behavior in the same type of congestion behaviors, and obtain the feature value range of the selected abnormal feature.

[0064] A transportation prediction and decision-making module, which is used to build a congestion prediction model to predict the congestion situation of transportation logs, and train based on the actual congestion situation of any transportation log; formulate corresponding adjustment decisions for various congestion behaviors based on the prediction results.

[0065] Among them, the transportation prediction and decision-making module includes a prediction model construction unit and an adjustment decision-making unit.

[0066] The prediction model construction unit is used to arbitrarily select a transportation log, analyze the deviation situation of the abnormal features included in any transportation area in the selected transportation log, and analyze the abnormal occurrence frequency of each abnormal feature to obtain the abnormal weight of each abnormal feature, and build a congestion prediction model to predict the congestion risk of each transportation area; obtain a risk prediction threshold for judging whether there is a congestion behavior in the transportation area through the actual congestion situation of each transportation log in each transportation area.

[0067] The adjustment decision-making unit is used to extract abnormal features from the transportation areas in any transportation log where the congestion risk value exceeds the risk prediction threshold, adjust the numerical values of the monitoring data corresponding to any abnormal feature, and generate an adjustment decision.

[0068] Among them, the prediction model construction unit includes:

[0069] Arbitrarily select a transportation area from the selected transportation log to obtain several abnormal features included in the transportation area, and obtain the feature value range of each abnormal feature. Among them, it is set that the feature value range of the jth abnormal feature in the transportation area is [d j (min), d j (max)]; obtain the normal value range (d1 j , d2 j ) of the jth abnormal feature, and set a feature deviation value Q for the jth abnormal feature j . If d j (max) < d1 j , then Q j = d j (max). If d2 j < d j (min), then Q j = d j (min); obtain the deviation amplitude of the jth abnormal feature as η j = [Min(|Q j - d1 j |, |Q j - d2 j |)] / [Select(d1 j , d2j )], where Min() is the minimum function and Select() is the selection function. If the minimum function is obtained, |Q j -d1 j |, then Select(d1 j ,d2 j )=d1 j , if we take the minimum function, we get |Q j -d2 j |, then Select(d1 j ,d2 j )=d2 j ;

[0070] Example 1: Assume that there is an abnormal feature in the conveying area, which is pressure abnormality, and the feature value range is (12, 15). The normal value range of the pressure feature is (5, 10). Therefore, the feature deviation value is 12, and the deviation amplitude is calculated to be η = (12-10) / 10 = 20%;

[0071] Get the number of occurrences of the jth abnormal feature in all transport logs as m j And the number of behaviors that contain the jth abnormal feature in all congestion behaviors is n j , calculate the abnormal weight f of the jth abnormal feature j =n j / m j ;

[0072] Construct a congestion prediction model to calculate the congestion risk prediction value R of the transportation area ex :

[0073]

[0074] If there is congestion in the selected transport area in the selected transport log, the congestion risk prediction value R ex Set as the abnormal prediction value; obtain the abnormal prediction value of the transportation area where any congestion behavior is located, and select the abnormal prediction value with the smallest value as the risk prediction threshold for judging whether there is congestion behavior in the transportation area;

[0075] Example 2: Obtain a transport area in the transport log, there are 3 abnormal features, and the deviation amplitude and abnormal weight are (20%, 0.8), (25%, 0.7), (20%, 0.9) respectively, and calculate the congestion risk prediction value R ex =0.2×0.8+0.25×0.7+0.2×0.9=0.16+0.175+0.18=0.515.

[0076] Among them, the adjustment decision-making unit includes:

[0077] Set the risk prediction threshold as R th , select a transportation area from any transportation log. If the congestion risk prediction value obtained in the selected transportation area is R ex , if R ex ≥R th , then extract all abnormal features included in the selected transportation area;

[0078] Obtain the monitoring device where any abnormal feature is located, formulate an adjustment plan to adjust the conveyor belt in the selected transportation area, so that the monitoring data range of the monitoring device is within the normal value range; formulate corresponding adjustment plans for all abnormal features to generate an adjustment decision for the selected transportation area.

[0079] Real-time prediction feedback module, used to monitor the real-time transportation process on the conveyor belt, obtain the real-time transportation log, perform congestion prediction on the real-time transportation log and make timely adjustments; monitor the change of abnormal features after adjustment, identify existing abnormal conditions and give reminders;

[0080] Among them, the real-time prediction feedback module includes a real-time transportation prediction unit and an abnormal congestion feedback unit;

[0081] Real-time output prediction unit, used to obtain the monitoring data of each monitoring device in the real-time transportation log, obtain the monitoring data of any monitoring device at the current monitoring time point. If the monitoring data at the current monitoring time point does not belong to the preset normal value range, extract abnormal features from the monitoring device and obtain the deviation amplitude η now and the abnormal weight f of the extracted abnormal features; obtain the abnormal weights of each abnormal feature and the deviation amplitudes of the corresponding monitoring data in any transportation area, and input them into the congestion prediction model to obtain the congestion risk prediction values of each transportation area in the real-time transportation log, and adjust the transportation areas where the congestion risk prediction values exceed the risk prediction threshold;

[0082] Abnormal congestion feedback unit, used to obtain the monitoring data of each monitoring device in the adjusted transportation area every unit time. If the monitoring data of one monitoring device does not belong to the normal value range in a continuous number of monitoring time points, send an abnormal reminder to the adjusted transportation area.

[0083] It is obvious to those skilled in the art that the present invention is not limited to the details of the above-described exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention. Therefore, in any regard, 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. Accordingly, all changes that fall within the meaning and scope of the equivalent elements of the claims are intended to be embraced within the present invention. Any reference signs in the claims should not be construed as limiting the claims involved.

Claims

1. A dynamic speed regulation intelligent matching system for a flexible production conveyor belt, characterized by: The matching system includes a transportation data analysis module, a congestion feature analysis module, a transportation prediction decision module and a real-time prediction feedback module; The transport data analysis module is used to collect data for each transport process of the conveyor belt through a preset monitoring device to generate a transport log; identify the congestion behavior existing in any transport log to obtain a congestion behavior set of any transport log; The congestion feature analysis module is used to compare the differences of each congestion behavior in the set of all congestion behaviors, classify different congestion behaviors into categories; identify abnormal features of any type of congestion behavior, and analyze the feature range of each abnormal feature; The transport prediction decision module is used to construct a congestion prediction model to predict the congestion situation of the transport log, and to perform training based on the actual congestion situation of any transport log; and to make corresponding adjustment decisions for various congestion behaviors based on the prediction results; The real-time prediction feedback module is used to monitor the real-time transportation process on the conveyor belt, obtain the real-time transportation log, predict congestion of the real-time transportation log and make timely adjustments; monitor the changes in abnormal characteristics after adjustment, identify existing abnormal situations and give reminders.

2. According to claim 1, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The transport data analysis module includes a transport log collection unit and a transport congestion identification unit; The conveying log collection unit is used to divide the conveyor belt into several conveying areas according to the preset functional type, and each conveying area is equipped with a corresponding monitoring device; each time the conveyor belt starts to convey, each monitoring device starts to collect conveying data for the corresponding conveying area until the conveyor belt stops conveying; the monitoring data collected during each complete conveying process is presented according to the monitoring time to obtain the corresponding conveying log; The transport congestion identification unit is used to arbitrarily select a transport log, arbitrarily select a monitoring device from the selected transport log, extract the monitoring data of the monitoring device at any monitoring time point, set the monitoring data monitored by each monitoring device as a type of monitoring data, and preset a normal value range for each type of monitoring data, wherein a normal value range r = (d1, d2) is preset for the selected monitoring device, wherein d1 is the minimum value, d2 is the maximum value, and the value of the monitoring data at the monitoring time point t0 is set to d t0 ,like Then obtain several monitoring time points that are adjacent and continuous to the monitoring time point t0. If the values ​​of the monitoring data at the several monitoring time points do not belong to the normal value range, then merge the several monitoring time points to obtain a time interval, and obtain the transportation behavior in the time interval to obtain a congestion behavior of the transportation log; obtain the congestion behavior of each time interval to obtain a congestion behavior set of the selected transportation log.

3. According to claim 2, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The congestion feature analysis module includes a transport congestion classification unit and an abnormal feature identification unit; The transport congestion classification unit is used to select any congestion behavior from a transport log, obtain the time interval and various monitoring data of the selected congestion behavior; compare the time interval and monitoring data between any two congestion behaviors, and determine whether to classify the two user behaviors; The abnormal feature recognition unit is used to analyze the abnormal situation of any congestion behavior in various monitoring data of any same type and extract a number of abnormal features; The monitoring data corresponding to any abnormal feature in different congestion behaviors are extracted to obtain the feature value range of any abnormal feature.

4. According to claim 3, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The transport congestion division unit comprises: Randomly select a transport log, randomly select a congestion behavior from the transport log, set the selected congestion behavior as the target congestion behavior, and obtain the time interval of the target congestion behavior as (t g1 ,t g2 ); Obtain the conveying rate of the conveyor belt in each conveying area, and obtain the expected time interval of each conveying area, wherein the expected time interval of the i-th conveying area is set as (t1 i ,t2 i ); if (t g1 ,t g2 )∈(t1 i ,t2 i ), the expected occurrence area of ​​the target congestion behavior is the i-th transport area, the actual occurrence area of ​​the target congestion behavior captured by the monitoring device is the a-th transport area, and a transport deviation flag F is set. If i≠a, the first deviation flag F is set (i,a) So that F = F (i,a) , if i = a, then set the second deviation flag F i So that F = F i ; Randomly select a monitoring device to present the changes of monitoring data over monitoring time, obtain the preset normal value range of the selected monitoring device, if there is a monitoring value at any monitoring time point in a continuous time interval that does not belong to the normal value range, then mark the selected monitoring device as abnormal, and obtain the set A of monitoring devices with abnormal marks in the target congestion behavior; Reselect a congestion behavior from the remaining congestion behaviors and set it as a comparison congestion behavior, and obtain a delivery deviation mark F of the comparison congestion behavior. ’ and the set A of all monitoring devices containing abnormal markers ’ , if F ’ =F and A = A ’ , the target congestion behavior and the comparison congestion behavior are classified as the same type of congestion behavior, otherwise, they are classified as different congestion behaviors.

5. According to claim 4, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The abnormal feature recognition unit comprises: Randomly select a group of similar congestion behaviors and extract the set of monitoring devices with abnormal marks in the same congestion behaviors; randomly select a monitoring device from the set of monitoring devices and obtain the monitoring data presented by the monitoring device in all transportation logs to obtain a complete value range (d min ,d max ), set the normal value range of the monitoring equipment to (d1, d2), if d1>d min , then we get a deviation value range p1=(d min ,d1), if d2<d max , then we get a deviation value range p2 = (d2, d max ); If the selected monitoring device has a deviation value range, the data type of the monitoring data is set to the abnormal characteristics of the selected monitoring device; the abnormal characteristics of each monitoring device in each transport log containing the same type of congestion behavior are extracted to obtain several abnormal characteristics of the same type of congestion behavior; Select an abnormal feature at random, obtain the transport log of any congestion behavior in the same congestion behavior, and obtain the deviation value range (d1 ’ ,d2 ’ ), if d1 ’ <d2<d2 ’ , then the actual deviation range of the transport log is (d2, d2 ’ ); extract the actual deviation range of each congestion behavior in the same congestion behavior and take the intersection to obtain the feature value range of the selected abnormal feature.

6. According to claim 5, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The transport prediction decision module includes a prediction model building unit and an adjustment decision making unit; The prediction model building unit is used to arbitrarily select a transport log, analyze the deviation of abnormal features contained in any transport area in the selected transport log, analyze the abnormal occurrence frequency of each abnormal feature, obtain the abnormal weight of each abnormal feature, and build a congestion prediction model to predict the congestion risk of each transport area; obtain the risk prediction threshold for judging whether there is congestion behavior in the transport area through the actual congestion situation of each transport log in each transport area; The adjustment decision making unit is used to extract abnormal features from any transportation area in the transportation log whose congestion risk value exceeds the risk prediction threshold, adjust the monitoring data corresponding to any abnormal feature, and generate an adjustment decision.

7. According to claim 6, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The prediction model building unit comprises: Select a transport area from the selected transport log, obtain several abnormal features contained in the transport area, and obtain the feature value range of each abnormal feature, wherein the feature value range of the jth abnormal feature in the transport area is set to [d j (min),d j (max)]; Get the normal value range of the jth abnormal feature (d1 j ,d2 j ), set a feature deviation value Q for the jth abnormal feature j , if d j (max)<d1 j , then Q j =d j (max), if d2 j <d j (min), then Q j =d j (min); the deviation amplitude of the j-th abnormal feature is η j =[Min(|Q j -d1 j |,|Q j -d2 j |)] / [Select(d1 j ,d2 j )], where Min() is the minimum function and Select() is the selection function. If the minimum function is obtained, |Q j -d1 j |, then Select(d1 j ,d2 j )=d1 j , if we take the minimum function, we get |Q j -d2 j |, then Select(d1 j ,d2 j )=d2 j ; Get the number of occurrences of the jth abnormal feature in all transport logs as m j And the number of behaviors that contain the jth abnormal feature in all congestion behaviors is n j , calculate the abnormal weight f of the jth abnormal feature j =n j / m j ; Construct a congestion prediction model to calculate the congestion risk prediction value R of the transportation area ex : If there is congestion in the selected transport area in the selected transport log, the congestion risk prediction value R ex Set as the abnormal prediction value; obtain the abnormal prediction value of the transportation area where any congestion behavior is located, and select the abnormal prediction value with the smallest value as the risk prediction threshold for judging whether there is congestion behavior in the transportation area.

8. According to claim 7, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The adjustment decision-making unit comprises: Set the risk prediction threshold as R th , select a transport area from any transport log, if the congestion risk prediction value obtained in the selected transport area is R ex , if R ex ≥R th , then all abnormal features contained in the selected conveying area are extracted; Obtain the monitoring device where any abnormal feature is located, formulate an adjustment plan to adjust the conveyor belt in the selected conveying area, so that the monitoring data range of the monitoring device is within the normal value range; formulate corresponding adjustment plans for all abnormal features, and generate an adjustment decision for the selected conveying area.

9. According to claim 8, a flexible production conveyor belt dynamic speed regulation intelligent matching system is characterized by: The real-time prediction feedback module includes a real-time transportation prediction unit and an abnormal congestion feedback unit; The real-time output prediction unit is used to obtain the monitoring data of each monitoring device in the real-time transmission log, and obtain the monitoring data of any monitoring device at the current monitoring time point. If the monitoring data at the current monitoring time point does not belong to the preset normal value range, abnormal features are extracted from the monitoring device, and the deviation amplitude η of the monitoring data is obtained. now and extracting the abnormal weight f of the abnormal feature; obtaining the abnormal weight of each abnormal feature in any transport area and the deviation amplitude of the corresponding monitoring data, and inputting them into the congestion prediction model to obtain the congestion risk prediction value of each transport area in the real-time transport log, and adjusting the transport area whose congestion risk prediction value exceeds the risk prediction threshold; The abnormal congestion feedback unit is used to obtain the monitoring data of each monitoring device in the adjusted transportation area every unit time. If the monitoring data of a monitoring device does not belong to the normal value range at several consecutive monitoring time points, an abnormal reminder is sent to the adjusted transportation area.

Images