Waste paper weighing data intelligent management method and system based on Internet of Things

Abstract

The embodiment of the invention provides a waste paper weighing data intelligent management method and system based on the Internet of Things, and the method comprises the steps: firstly obtaining a real-time weighing data set uploaded by a plurality of waste paper collection devices, then carrying out the dynamic correction of the real-time weighing data set, generating a standardized weighing data set, extracting the weight change trend characteristics, and carrying out the dynamic correction of the real-time weighing data set; abnormality analysis is performed on the weight change trend characteristics, the operation abnormality type and the corresponding abnormality confidence coefficient of the waste paper collection equipment are determined, an equipment maintenance instruction set is generated according to the operation abnormality type and the corresponding abnormality confidence coefficient and sent to target equipment to trigger calibration operation, and finally a periodic data management report is generated based on calibrated data. Intelligent management of the weighing data of the waste paper collecting equipment is realized, and the data accuracy and the equipment maintenance efficiency are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Things, in particular to a waste paper weighing data intelligent management method and system based on Internet of Things. BACKGROUND

[0002] In the waste paper recycling industry, waste paper collection equipment as an important front-end equipment in the waste paper recycling process, its running state and data accuracy directly affect the subsequent waste paper processing, classification and recycling efficiency. Traditionally, the weighing data of waste paper collection equipment mainly relies on manual recording or simple data transmission system, which is not only inefficient, but also prone to inaccurate data due to human factors or equipment errors. In addition, due to the lack of real-time monitoring and analysis of the running state of waste paper collection equipment, it is often difficult to find and handle the equipment failure or abnormality in time, thereby affecting the stability and efficiency of the entire waste paper recycling system. With the rapid development of Internet of Things technology, how to use Internet of Things technology to realize the intelligent management of waste paper collection equipment weighing data, improve data accuracy and equipment maintenance efficiency, has become a problem to be solved in the current waste paper recycling industry. SUMMARY

[0003] Therefore, the purpose of the embodiments of the present application is to provide a waste paper weighing data intelligent management method and system based on Internet of Things.

[0004] According to one aspect of the embodiments of the present application, a waste paper weighing data intelligent management method based on Internet of Things is provided, the method comprising: obtaining a set of real-time weighing data uploaded by a plurality of waste paper collection equipment, the set of real-time weighing data containing device identifiers and weight measurement values at different time points; performing dynamic correction processing on the set of real-time weighing data to generate a set of standardized weighing data and extract weight change trend characteristics corresponding to each device identifier; performing abnormality analysis on the weight change trend characteristics to determine the running abnormality type of the waste paper collection equipment and the corresponding abnormality confidence; generating a set of equipment maintenance instructions according to the running abnormality type and the abnormality confidence, and sending the set of equipment maintenance instructions to the target waste paper collection equipment to trigger calibration operation; generating a periodic data management report based on the set of standardized weighing data after calibration.

[0005] In an aspect of the embodiments of the present application, a system for intelligent management of waste paper weighing data based on Internet of Things is provided, which comprises a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory can communicate with each other through the communication bus; the memory is used for storing a computer program; and the processor is used for executing the computer program to implement the steps of the method for intelligent management of waste paper weighing data based on Internet of Things.

[0006] In another aspect of the embodiments of the present application, a readable storage medium is provided, and the readable storage medium stores a computer program, which can execute the steps of the method for intelligent management of waste paper weighing data based on Internet of Things when the computer program is run by a processor.

[0007] Through any one of the above aspects, the comprehensive and dynamic management of real-time weighing data uploaded by multiple waste paper collection devices is realized through the integration of Internet of Things technology, which not only can dynamically correct the collected data to generate a standardized weighing data set, thereby significantly improving the accuracy and reliability of the data, but also can accurately determine the operation abnormal type and the corresponding abnormal confidence of the waste paper collection device by extracting the weight change trend characteristics corresponding to each device identifier and performing abnormal analysis. Based on these analysis results, a device maintenance instruction set can be automatically generated and sent to the target waste paper collection device to trigger a calibration operation, thereby realizing the remote monitoring and intelligent maintenance of the device, greatly reducing the need for manual intervention and improving the timeliness and effectiveness of device maintenance. Finally, a periodic data management report is generated based on the standardized weighing data set after calibration.

[0008] In order to make the above-mentioned purposes, features and advantages of the embodiments of the present application more obvious and easy to understand, the following will be described in detail in combination with the embodiments and the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0009] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0010] Figure 1 Fig. 1 shows a component schematic diagram of the system for intelligent management of waste paper weighing data based on Internet of Things provided by the embodiments of the present application; Figure 2 Fig. 2 shows a flow schematic diagram of the method for intelligent management of waste paper weighing data based on Internet of Things provided by the embodiments of the present application. DETAILED DESCRIPTION

[0011] In order to better understand the technical scheme of the present application, the technical scheme in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. According to the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0012] The terms "first", "second", "third", and so on, if any, in the description, claims, and drawings of the present application (if any) are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device including a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0013] Figure 1 An exemplary component diagram of the Internet of Things based waste paper weighing data intelligent management system 100 is shown. The Internet of Things based waste paper weighing data intelligent management system 100 can include one or more processors 104, such as one or more central processing units (CPUs), each of which can implement one or more hardware threads. The Internet of Things based waste paper weighing data intelligent management system 100 can also include any storage medium 106 for storing any kind of information such as code, settings, data, etc. Without limitation, for example, the storage medium 106 can include any one or a combination of more than one of the following: any type of RAM, any type of ROM, a flash memory device, a hard disk, an optical disk, etc. More generally, any storage medium can store information using any technology. Further, any storage medium can provide volatile or non-volatile retention of information. Further, any storage medium can represent a fixed or removable component of the Internet of Things based waste paper weighing data intelligent management system 100. In one case, the Internet of Things based waste paper weighing data intelligent management system 100 can perform any operation associated with the relevant instructions stored in any storage medium or combination of storage media when the processor 104 executes the instructions. The Internet of Things based waste paper weighing data intelligent management system 100 also includes one or more drive units 108, such as a hard disk drive unit, an optical disk drive unit, etc., for interacting with any storage medium.

[0014] The Internet of Things based waste paper weighing data intelligent management system 100 also includes an input / output 110 (I / O) for receiving various inputs (via input units 112) and for providing various outputs (via output units 114). One particular output mechanism can include a presentation device 116 and a presence-dependent graphical user interface (GUI) 118. The Internet of Things based waste paper weighing data intelligent management system 100 can also include one or more network interfaces 120 for exchanging data with other devices via one or more communication units 122. One or more communication buses 124 couple the above-described components together.

[0015] The communication units 122 can be implemented in any way, for example, through a local area network, a wide area network (e.g., the Internet), a point-to-point connection, etc., or any combination thereof. The communication units 122 can include any combination of hardwired links, wireless links, routers, gateway functionality, name-based Internet of Things based waste paper weighing data intelligent management system 100, etc., governed by any protocol or combination of protocols.

[0016] Figure 2 A flowchart of the Internet of Things based waste paper weighing data intelligent management method and system provided by the embodiments of the present application is shown. The Internet of Things based waste paper weighing data intelligent management method and system can be executed by the Internet of Things based waste paper weighing data intelligent management system 100 shown in Figure 1 The detailed steps of the Internet of Things based waste paper weighing data intelligent management method are introduced as follows.

[0017] Step S110: Obtain a real-time weighing data set uploaded by a plurality of waste paper collection devices, the real-time weighing data set containing device identifiers and weight measurement values at different time points.

[0018] In the actual scenario of waste paper recycling, there are a plurality of waste paper collection devices distributed at different locations. These devices have weighing functions and will measure the weight of the collected waste paper at set time intervals and upload the measured data. Each uploaded data unit contains a device identifier and a weight measurement value as well as corresponding time point information. The device identifier is used to uniquely identify each waste paper collection device, and different devices have different identifiers, such as device a, device b, etc. The weight measurement value is the weight of the waste paper measured by the device at a particular time point. As time goes on, multiple devices continuously upload data, thereby forming a real-time weighing data set. The real-time weighing data set can be represented as S={(ID1, t1, m1), (ID2, t2, m2), …, (IDn, tn, mn)}, where IDi represents the identifier of the ith device, ti represents the time point, and mi represents the weight measurement value of the device at the time point.

[0019] Step S120: dynamically correcting the real-time weighing data set, generating a standardized weighing data set and extracting the weight change trend characteristics corresponding to each device identifier.

[0020] Since the real-time weighing data can be disturbed by various factors, such as environmental factors, minor faults of the device itself, etc., in order to ensure the accuracy and reliability of the data, dynamic correction processing is needed for the real-time weighing data set. The specific operation is as follows: Step S121: identifying the measurement value sequence in the continuous time window in the real-time weighing data set, and calculating the measurement value difference degree of adjacent time points.

[0021] In the real-time weighing data set, a continuous time window is selected. The size of the time window can be set according to actual needs, for example, set as a fixed time period. For each device, there will be a series of measurement values in the time window. Taking device a as an example, in the time window [t_start, t_end], its measurement value sequence can be represented as Ma={m_a1, m_a2, …, m_ak}, where k is the number of measurements in the time window. In order to analyze the change of the measurement value, the measurement value difference degree of adjacent time points needs to be calculated. For adjacent measurement values m_ai and m_a(i+1), the measurement value difference degree Am_ai can be calculated by m_a(i+1)-m_ai. In this way, for the measurement value sequence of device a in the time window, a series of measurement value difference degrees {Am_a1, Am_a2, …, Am_a(k-1)} can be obtained.

[0022] Step S122: when the measurement value difference degree exceeds the preset fluctuation threshold, obtaining the environmental parameter set of the corresponding waste paper collection device, the environmental parameter set including temperature measurement value, humidity measurement value and device vibration intensity.

[0023] A fluctuation threshold T is set in advance to determine whether the change of the measurement value is abnormal. When the calculated measurement value difference degree Am_ai exceeds the fluctuation threshold T, it means that the change of the measurement value may be affected by external environmental factors. At this time, the environmental parameter set of the corresponding waste paper collection device needs to be obtained. The environmental parameter set E includes the temperature measurement value T_meas, the humidity measurement value H_meas and the device vibration intensity V_meas. These environmental parameters can be obtained by temperature sensors, humidity sensors and vibration sensors installed near the device. For example, for device a, when a certain measurement value difference degree Am_aj exceeds the fluctuation threshold T, the environmental parameter set E_a={T_a, H_a, V_a} of the device at this time is obtained, where T_a is the temperature measurement value, H_a is the humidity measurement value, and V_a is the device vibration intensity.

[0024] Step S123: calling a dynamic compensation model to perform weight allocation on the set of environmental parameters to generate an environmental interference correction coefficient.

[0025] The dynamic compensation model is a pre-trained model for performing weight allocation on the set of environmental parameters. The dynamic compensation model can be generated in the following steps to generate the environmental interference correction coefficient: Step S1231: inputting the temperature measurement value, humidity measurement value, and equipment vibration intensity into a pre-trained feature mapping layer to generate an environmental interference vector.

[0026] The pre-trained feature mapping layer is part of the dynamic compensation model, which maps the input temperature measurement value T_meas, humidity measurement value H_meas, and equipment vibration intensity V_meas to generate an environmental interference vector. The feature mapping layer can adopt a neural network structure, such as a multi-layer perceptron (MLP). The set of environmental parameters E = {T_meas, H_meas, V_meas} is input into the feature mapping layer, and an environmental interference vector V_env = [v1, v2, v3] is output, where v1, v2, and v3 are the characteristic values of the mapped temperature, humidity, and equipment vibration intensity, respectively.

[0027] Step S1232: performing multi-order nonlinear transformation on the environmental interference vector to obtain a dynamic influence weight corresponding to each environmental parameter.

[0028] In order to determine the influence degree of each environmental parameter on the measurement value, multi-order nonlinear transformation is needed on the environmental interference vector. A series of nonlinear functions, such as ReLU function, Sigmoid function, etc., can be used to transform the environmental interference vector multiple times. After multi-order nonlinear transformation, a dynamic influence weight corresponding to each environmental parameter is obtained. Let the dynamic influence weight vector be W = [w1, w2, w3], where w1, w2, and w3 are the dynamic influence weights of temperature, humidity, and equipment vibration intensity, respectively.

[0029] Step S1233: performing weighted fusion on the set of environmental parameters according to the dynamic influence weight to generate the environmental interference correction coefficient.

[0030] The dynamic influence weight vector W and the set of environmental parameters E are weighted and fused to obtain the environmental interference correction coefficient C. The specific calculation method is: C = w1*T_meas + w2*H_meas + w3*V_meas. The environmental interference correction coefficient reflects the comprehensive influence degree of environmental factors on the measurement value.

[0031] Step S124: After compensating the measurement values in the real-time weighing data set according to the environmental interference correction coefficient and obtaining the standardized weighing data set S' and marking the time point of the corrected measurement value source, the weight change trend characteristics corresponding to each device identifier are extracted.

[0032] The measurement values in the real-time weighing data set are compensated by using the calculated environmental interference correction coefficient C. For the measurement value m_a(t) of device a at time point t, the compensated measurement value m_a'(t) can be calculated by m_a(t)-C. After such compensation calculation is performed on all measurement values in the real-time weighing data set, the standardized weighing data set S' can be obtained. At the same time, in order to facilitate subsequent analysis, it is necessary to mark the time point of the corrected measurement value source. Then, for each device identifier, the weight change trend characteristics corresponding thereto are extracted from the standardized weighing data set. The weight change trend characteristics can be obtained by analyzing the compensated measurement values in a period of time, such as calculating the change rate, slope, etc. of the measurement values. For example, for device a, the weight change rate R_a in a period of time can be calculated, which is calculated by subtracting the compensated measurement values of adjacent time points and then dividing by the time interval.

[0033] Step S130: Abnormal analysis is performed on the weight change trend characteristics to determine the operation abnormality type of the waste paper collecting device and the corresponding abnormal confidence.

[0034] In order to ensure the normal operation of the waste paper collecting device, abnormal analysis needs to be performed on the extracted weight change trend characteristics, and the specific operation is as follows: Step S131: Periodic fluctuation patterns and sudden offset amounts in the weight change trend characteristics are extracted.

[0035] In the weight change trend characteristics, there are periodic fluctuation patterns and sudden offset amounts. The periodic fluctuation pattern refers to the periodic change of the weight measurement value in a set time, such as the weight change rule in a fixed period of time, e.g. daily, weekly, etc. The periodic fluctuation pattern can be extracted by performing frequency spectrum analysis, etc. on the weight change trend characteristics in a period of time. The sudden offset amount refers to the sudden change of the weight measurement value that deviates from the normal range. A threshold value can be set, and when the weight change rate exceeds the threshold value, it is considered that a sudden offset amount has occurred.

[0036] Step S132: Similarity matching is performed between the periodic fluctuation pattern and the normal fluctuation range in the historical data to generate a periodic abnormality score.

[0037] The normal fluctuation range in the historical data is pre-stored, which can be expressed as an interval [min_fluct, max_fluct]. The extracted periodic fluctuation pattern is matched with the normal fluctuation range in the historical data in terms of similarity. Similarity calculation methods such as Euclidean distance, cosine similarity, etc. can be used to calculate the similarity of the periodic fluctuation pattern and the normal fluctuation range. According to the similarity calculation result, a periodic anomaly score is generated. If the similarity is high, it means that the periodic fluctuation pattern is within the normal range, and the periodic anomaly score is low; on the contrary, if the similarity is low, it means that the periodic fluctuation pattern may be abnormal, and the periodic anomaly score is high.

[0038] Step S133: Time decay coefficient calculation is performed on the abrupt offset to generate an abrupt anomaly score.

[0039] For the abrupt offset, the time factor of its occurrence needs to be considered. The specific steps are as follows: Step S1331: The starting time point and the duration length corresponding to the abrupt offset are identified.

[0040] When the abrupt offset is detected, its corresponding starting time point t_start and duration length Δt are recorded. The starting time point represents the time when the abrupt offset starts to appear, and the duration length represents the time during which the abrupt offset lasts.

[0041] Step S1332: A time decay coefficient is determined according to the duration length and a preset decay function, and the decay function is an exponential function that decreases with the increase of the time length.

[0042] A decay function f(Δt) is pre-set, which is an exponential function that decreases with the increase of the time length, for example, f(Δt)=e^(-λ*Δt), where λ is a decay coefficient. According to the duration length Δt of the abrupt offset, the decay function is substituted to obtain the time decay coefficient α.

[0043] Step S1333: The time decay coefficient is multiplied by the absolute value of the abrupt offset to obtain the abrupt anomaly score; wherein when the cumulative value of the abrupt anomaly score exceeds a preset abrupt threshold, it is marked as a device hardware failure type.

[0044] The time decay coefficient α is multiplied by the absolute value |ΔS| of the abrupt offset to obtain the abrupt anomaly score S_abrupt=α*|ΔS|. At the same time, the abrupt anomaly score is accumulated, and if the cumulative value exceeds a preset abrupt threshold T_abrupt, the device is marked as a device hardware failure type.

[0045] Step S134: superimpose the periodic anomaly score and the abrupt anomaly score to obtain a comprehensive anomaly score.

[0046] The calculated periodic anomaly score S_periodic and the abrupt anomaly score S_abrupt are superimposed to obtain a comprehensive anomaly score S_total=S_periodic+S_abrupt. The comprehensive anomaly score reflects the comprehensive abnormality degree of the device in terms of periodic fluctuations and abrupt deviations.

[0047] Step S135: determine the running abnormality type and the corresponding abnormality confidence according to the comprehensive anomaly score and a preset anomaly level division rule.

[0048] The anomaly level division rule is preset, which can divide the anomaly into different levels such as mild anomaly, moderate anomaly, and severe anomaly according to the range of the comprehensive anomaly score. According to the calculated comprehensive anomaly score S_total, the running abnormality type of the device is determined according to the anomaly level division rule. At the same time, according to the corresponding relationship between the anomaly level and the abnormality confidence, the corresponding abnormality confidence is determined. For example, the abnormality confidence corresponding to the mild anomaly is low, and the abnormality confidence corresponding to the severe anomaly is high.

[0049] Step S140: generate a device maintenance instruction set according to the running abnormality type and the abnormality confidence, and send the device maintenance instruction set to the target waste paper collection device to trigger a calibration operation.

[0050] According to the determined running abnormality type and abnormality confidence, a corresponding device maintenance instruction set is generated, and the specific operation is as follows: Step S141: match the preset maintenance strategy library according to the running abnormality type to obtain the maintenance operation sequence and operation priority corresponding to the running abnormality type.

[0051] A maintenance strategy library is pre-established, which stores the maintenance operation sequence and operation priority corresponding to different running abnormality types. According to the determined running abnormality type, matching is performed in the maintenance strategy library to find the corresponding maintenance operation sequence and operation priority. For example, for a mild anomaly, the maintenance operation sequence may include a simple calibration operation; for a severe anomaly, the maintenance operation sequence may include more complex device repair and component replacement operations.

[0052] Step S142: dynamically adjust the execution order of the maintenance operation sequence based on the abnormality confidence to generate an optimized maintenance operation sequence.

[0053] The abnormal confidence reflects the possibility and severity of the abnormality. According to the abnormal confidence, the execution order of the maintenance operation sequence is dynamically adjusted. If the abnormal confidence is high, the maintenance operation that is more critical to solving the abnormality is preferentially executed; if the abnormal confidence is low, the operation order can be appropriately adjusted, and some simple check operation is executed first. Through such dynamic adjustment, an optimized maintenance operation sequence is generated.

[0054] Step S143: binding the optimized maintenance operation sequence with the identifier of the target waste paper collecting device, to generate the device maintenance instruction set.

[0055] Binding the optimized maintenance operation sequence with the identifier of the target waste paper collecting device ensures that each maintenance operation corresponds to a specific device. In this way, a device maintenance instruction set is generated, which can be represented as I={(ID1, OP1), (ID2, OP2), …, (IDn, OPn)}, where IDi represents the device identifier, and OPi represents the optimized maintenance operation sequence corresponding to the device.

[0056] Step S144: sending the device maintenance instruction set sorted according to the operation priority to the target device, and recording the instruction execution state.

[0057] The device maintenance instruction set is sorted according to the operation priority, with the operation of high priority arranged in front. Then the sorted device maintenance instruction set is sent to the target waste paper collecting device to trigger the calibration operation. At the same time, the execution state of each instruction is recorded, including whether the instruction is sent successfully, whether the device receives and executes the instruction, and other information.

[0058] Step S1441: receiving the instruction response data returned by the target waste paper collecting device, the instruction response data containing the maintenance operation completion time and the measurement value after operation.

[0059] After receiving the maintenance instruction, the target waste paper collecting device will execute the corresponding maintenance operation and return the instruction response data. The instruction response data contains the maintenance operation completion time t_finish and the measurement value m_post after operation. The maintenance operation completion time represents the time when the device completes the maintenance operation, and the measurement value after operation represents the weight of waste paper measured by the device after completing the maintenance operation.

[0060] Step S1442: comparing the maintenance operation completion time with the preset standard maintenance time to generate a maintenance efficiency score.

[0061] A standard maintenance duration T standard is preset. The maintenance operation completion time t finish is compared with the standard maintenance duration T standard. If the maintenance operation completion time is less than the standard maintenance duration, it indicates that the maintenance efficiency is high, and the maintenance efficiency score is high; otherwise, if the maintenance operation completion time is greater than the standard maintenance duration, it indicates that the maintenance efficiency is low, and the maintenance efficiency score is low. A score function can be set to generate the maintenance efficiency score S efficiency according to the comparison result.

[0062] Step S1443: generating a maintenance effect score according to the deviation value of the post-operation measurement value from the standardized weighing data set.

[0063] The deviation value Δm = |m post-m standard | of the post-operation measurement value m post and the measurement value m standard of the corresponding time point in the standardized weighing data set is calculated. The maintenance effect score S effect is generated according to the size of the deviation value. The smaller the deviation value, the better the maintenance effect, and the higher the maintenance effect score; the larger the deviation value, the worse the maintenance effect, and the lower the maintenance effect score.

[0064] Step S1444: storing the maintenance efficiency score and the maintenance effect score in association to the maintenance record database, and updating the operation priority in the maintenance strategy library.

[0065] The maintenance efficiency score S efficiency and the maintenance effect score S effect are stored in association to the maintenance record database for subsequent analysis and query. At the same time, the operation priority in the maintenance strategy library is updated according to the maintenance efficiency score and the maintenance effect score. If the efficiency and effect of a certain maintenance operation are both good, the priority thereof can be appropriately increased; otherwise, if the efficiency and effect of a certain maintenance operation are poor, the priority thereof can be appropriately decreased.

[0066] Step S150: generating a periodic data management report based on the calibrated standardized weighing data set.

[0067] In order to comprehensively manage and analyze the waste paper recycling situation, a periodic data management report is generated based on the calibrated standardized weighing data set, and the specific operation is as follows: Step S151: counting the abnormal occurrence frequency and maintenance response time distribution of all waste paper collection devices within a preset period.

[0068] A preset period is set, such as a week, a month, etc. In the preset period, the frequency of occurrence of abnormalities of all waste paper collection devices is counted. The frequency of occurrence of abnormalities can be obtained by calculating the number of times of occurrence of abnormalities of each device in the period. At the same time, the distribution of maintenance response time is counted, and the maintenance response time refers to the time interval from the occurrence of an abnormality of a device to the start of the maintenance operation. The maintenance response time can be divided into different intervals, and the number of maintenance responses in each interval is counted, so as to obtain the distribution of maintenance response time.

[0069] Step S152: extracting the weight cumulative value in the calibrated normalized weighing data set to generate a regional waste paper recycling amount distribution map.

[0070] Step S1521: obtaining the geographic location information and the home area identifier of each waste paper collection device.

[0071] Each waste paper collection device has its corresponding geographic location information, such as latitude and longitude coordinates, etc. At the same time, each device is assigned a home area identifier, which represents the area to which the device belongs. These information can be obtained and managed through a geographic information system (GIS).

[0072] Step S1522: grouping and aggregating the weight cumulative value according to the home area identifier to generate the total amount of waste paper recycling in each area.

[0073] According to the home area identifier, the weight cumulative value in the calibrated normalized weighing data set is grouped. For each area, the weight cumulative values of all devices in the area are summed to obtain the total amount of waste paper recycling in the area. For example, there are devices a1, a2, a3 in area A, and their weight cumulative values are M_a1, M_a2, M_a3, respectively. Then the total amount of waste paper recycling in area A is M_A=M_a1+M_a2+M_a3.

[0074] Step S1523: normalizing the total amount of waste paper recycling with the regional population density data to generate a unit population recycling efficiency value.

[0075] The population density data D of each area is obtained. In order to eliminate the influence of the difference in population number of different areas, the total amount of waste paper recycling is normalized with the regional population density data. Normalization formula can be used, such as dividing the total amount of waste paper recycling by the population number of the area to obtain the unit population recycling efficiency value E=M / (D*S), where M is the total amount of waste paper recycling, and S is the area.

[0076] Step S1524: drawing a heat distribution map based on the unit population recycling efficiency value, and marking the area with an efficiency value lower than a preset threshold as an area to be optimized.

[0077] According to the unit population recovery efficiency value, a heat distribution map is drawn. The heat distribution map can intuitively show the waste paper recovery efficiency of each region. A preset threshold T_efficiency is set, and the regions with unit population recovery efficiency values lower than the threshold are marked as optimization areas. These regions may have lower waste paper recovery efficiency problems and need further analysis and improvement.

[0078] Step S153: generating a device reliability evaluation index according to the abnormality occurrence frequency and the maintenance response time distribution.

[0079] Considering the abnormality occurrence frequency and the maintenance response time distribution, a device reliability evaluation index is generated. A weighted sum method can be used to assign different weights to the abnormality occurrence frequency and the maintenance response time distribution, and then the two are weighted and summed to obtain the device reliability evaluation index R. For example, let the weight of the abnormality occurrence frequency be w1 and the weight of the maintenance response time distribution be w2, then the device reliability evaluation index R = w1*F + w2*T_dist, where F is the abnormality occurrence frequency and T_dist is a quantitative value of the maintenance response time distribution.

[0080] Step S154: correlating the regional waste paper recovery amount distribution map with the device reliability evaluation index to generate the data management report containing optimization suggestions.

[0081] After obtaining the regional waste paper recovery amount distribution map and the device reliability evaluation index, correlation analysis is needed to comprehensively understand the operation of the waste paper recovery system, and to generate a data management report containing optimization suggestions based on the analysis results. The purpose of correlation analysis is to find the potential relationship between regional waste paper recovery amount and device reliability, so as to provide basis for system optimization.

[0082] First, each region in the regional waste paper recovery amount distribution map is matched with the corresponding device reliability evaluation index. The waste paper recovery amount of each region can be represented by a vector, denoted as R = {R1, R2, …, Rn}, where Ri represents the waste paper recovery amount of the i-th region. The device reliability evaluation index can also be represented by a vector, denoted as S = {S1, S2, …, Sn}, where Si represents the device reliability evaluation index corresponding to the i-th region.

[0083] For the correlation analysis, the method of correlation analysis can be used. The correlation coefficient between the waste paper recycling amount of each region and the equipment reliability evaluation index is calculated. The correlation coefficient can measure the strength of the linear relationship between two variables. Statistical methods can be used to calculate the correlation coefficient, such as Pearson correlation coefficient. The correlation coefficient matrix C is calculated, where Cij represents the correlation coefficient between the waste paper recycling amount of the ith region and the equipment reliability evaluation index of the jth region.

[0084] According to the correlation coefficient matrix C, the relationship between the waste paper recycling amount and the equipment reliability of the region can be further analyzed. If the correlation coefficient between the waste paper recycling amount and the equipment reliability evaluation index of a certain region is high, it indicates that the waste paper recycling situation of the region may be closely related to the reliability of the equipment. For example, if the correlation coefficient is positive and close to 1, it indicates that the higher the equipment reliability, the higher the waste paper recycling amount may be; if the correlation coefficient is negative and close to -1, it indicates that the lower the equipment reliability, the higher the waste paper recycling amount may be, which may imply that there are other factors affecting the waste paper recycling amount, such as the large amount of waste paper generated in the region.

[0085] After analyzing the correlation, combining the information of the region to be optimized in the waste paper recycling amount distribution map and the equipment reliability evaluation index, the optimization suggestion is generated. For the region with low equipment reliability and low waste paper recycling amount, it can be suggested to strengthen the maintenance and management of the equipment, improve the reliability of the equipment, and thus possibly improve the waste paper recycling amount. For example, the inspection frequency of the equipment can be increased, and the aging parts can be replaced in time. For the region with high equipment reliability but low waste paper recycling amount, it may be necessary to further analyze the waste paper generation situation, recycling channel and other factors in the region, and suggest optimizing the recycling channel to improve the recycling efficiency.

[0086] The analysis results and optimization suggestions are arranged into a data management report. The data management report should include the waste paper recycling amount distribution map of the region, the equipment reliability evaluation index, the correlation analysis results and specific optimization suggestions, etc. The structure of the report can be divided into introduction, data overview, correlation analysis results, optimization suggestions and conclusion, etc.

[0087] Step S155: Push the data management report to the specified terminal and trigger the visual display operation.

[0088] After the generation of the data management report, it needs to be pushed to the specified terminal so that the relevant personnel can obtain and view the report content in time. The specified terminal can be the computer, mobile phone and other equipment of the management personnel.

[0089] First, the information of the designated terminal is determined, including the type of the terminal, the device identifier, etc. These information can be obtained through the terminal information library pre-set in the system. Then, the appropriate push method is selected according to the type of the terminal. For computer terminals, data management reports can be pushed through email, internal office systems, etc.; for mobile terminals, reports can be pushed through SMS, mobile applications, etc.

[0090] At the same time of pushing the data management report, the visualization operation is triggered. Visualization can present the data and information in the report in the form of intuitive charts, graphs, etc., making it easy for relevant personnel to understand and analyze. Professional visualization tools such as Tableau, PowerBI, etc. can be used to visualize the regional waste paper recycling amount distribution chart, equipment reliability evaluation index, etc. in the report. For example, the regional waste paper recycling amount distribution chart is displayed in the form of a map, and the waste paper recycling amount of different regions is represented by different colors or sizes; the equipment reliability evaluation index is displayed in the form of a column chart or a line chart, which intuitively shows the differences in equipment reliability in different regions.

[0091] After receiving the data management report on the designated terminal and performing visualization, relevant personnel can adjust and optimize the waste paper recycling system according to the content and optimization suggestions in the report, thereby improving the efficiency of waste paper recycling and the reliability of the equipment.

[0092] Next, the training process of the dynamic compensation model is introduced in detail, which is the key link for dynamic correction of real-time weighing data in the entire method.

[0093] Step S210: Prepare training data.

[0094] The quality and diversity of training data are crucial to the performance of the dynamic compensation model. First, a large amount of environmental parameter data and corresponding weighing data are collected. Environmental parameter data includes temperature measurement, humidity measurement and equipment vibration intensity, and weighing data is the measurement value of the waste paper collection equipment under corresponding environmental conditions.

[0095] The collected data is divided into training set, validation set and test set according to the set proportion. The training set is used for model training, the validation set is used for adjusting the parameters of the model during training, and the test set is used for evaluating the final performance of the model.

[0096] To ensure the quality of the data, the collected data is pre-processed. The pre-processing includes data cleaning, normalization, etc. Data cleaning is to remove noise, outliers, etc. in the data, to ensure the accuracy and consistency of the data. Normalization is to convert the environmental parameter data and weighing data of different ranges to the same scale, to avoid the influence of data scale difference on model training. For example, the min-max normalization method can be used to map the data of each parameter to the [0, 1] interval.

[0097] Step S220: Constructing a dynamic compensation model.

[0098] The dynamic compensation model adopts a neural network structure, including an input layer, a feature mapping layer, a multi-order nonlinear transformation layer, and an output layer.

[0099] The input layer receives environmental parameter data, i.e. temperature measurement, humidity measurement, and equipment vibration intensity. The number of neurons in the input layer is equal to the dimension of the environmental parameters, i.e. 3.

[0100] The feature mapping layer is the key part of the model, which maps the input environmental parameter data to generate an environmental interference vector. The feature mapping layer can adopt a multi-layer perceptron (MLP) structure, containing multiple hidden layers. Each hidden layer is composed of multiple neurons, and the neurons are connected through weights. The number of neurons and the number of layers in the hidden layer can be adjusted according to the actual situation to achieve the best mapping effect.

[0101] The multi-order nonlinear transformation layer performs multiple nonlinear transformations on the environmental interference vector to obtain the dynamic influence weight corresponding to each environmental parameter. This multi-order nonlinear transformation layer can use a series of nonlinear functions, such as ReLU function, Sigmoid function, etc. Through multiple nonlinear transformations, the complex relationship between environmental parameters can be captured.

[0102] The output layer outputs the environmental interference correction coefficient. The number of neurons in the output layer is 1.

[0103] Step S230: Training the dynamic compensation model.

[0104] When training the dynamic compensation model, the training set data is used for training. First, the parameters of the model are initialized, including the weights and biases between the neurons of each layer.

[0105] The backpropagation algorithm is used for model training. The backpropagation algorithm calculates the error between the model output and the true value, then propagates the error back to each layer of the model, adjusts the weights and biases between the neurons to reduce the error.

[0106] During the training process, the model is validated using the validation set data. Every set number of training steps, the validation set data is input into the model, and the error of the model on the validation set is calculated. According to the error on the validation set, the training parameters of the model, such as the learning rate, are adjusted to avoid overfitting or underfitting of the model.

[0107] The goal of training is to minimize the error of the model on the training set and the validation set. Mean squared error (MSE) can be used as a loss function to measure the difference between the model's output of the environmental disturbance correction coefficient and the true environmental disturbance correction coefficient.

[0108] Step S240: Evaluate the dynamic compensation model.

[0109] The trained dynamic compensation model is evaluated using the test set data. The test set data is input into the model to obtain the model's output of the environmental disturbance correction coefficient. The error of the model on the test set is calculated, such as mean squared error, mean absolute error, etc.

[0110] According to the evaluation results, it is judged whether the performance of the model meets the requirements. If the error of the model is within an acceptable range, it means that the performance of the model is good and can be used for actual dynamic correction processing; if the error of the model is large, the model needs to be adjusted, such as increasing the number of neurons in the hidden layer, adjusting the learning rate, etc., and then retraining and evaluating.

[0111] In addition, during the entire data processing and management process, the protection of privacy-sensitive data is involved. For example, the geographic location information of the waste paper collection device is privacy-sensitive data, and appropriate privacy protection and anti-leakage technical means need to be taken.

[0112] For geographic location information, encryption technology can be used for protection. In the data collection stage, symmetric encryption algorithm is used to encrypt the geographic location information, and the encrypted data is stored in the system. In the data transmission process, a secure transmission protocol such as SSL / TLS protocol is used to ensure the security of the data. In the data use stage, only authorized personnel can decrypt and use the geographic location information.

[0113] At the same time, a strict access control mechanism is established. Different access permissions are set for the data in the system, and only personnel with corresponding permissions can access and process privacy-sensitive data. Regular security audits are conducted on the system to check whether there is a risk of data leakage, and timely measures are taken for prevention.

[0114] Through the above detailed steps and technical means, the method can realize intelligent management of waste paper weighing data, including data acquisition, correction, abnormality analysis, maintenance instruction generation, and data report generation, etc. functions, while ensuring the security of the data and the accuracy of the model, providing strong support for the efficient operation of the waste paper recycling system.

[0115] In practical applications, the dynamic compensation model and the maintenance strategy library can be continuously optimized over time and with the accumulation of data. For example, new environmental parameter data and weighing data are collected regularly, and the dynamic compensation model is retrained to improve the adaptability and accuracy of the model. According to the data in the maintenance record database, the effects and efficiencies of different maintenance operations are analyzed, and the operation priorities and maintenance operation sequences in the maintenance strategy library are further adjusted.

[0116] In addition, the method can also be integrated with other systems, such as waste paper recycling scheduling systems, logistics management systems, etc. Through integration with these systems, the overall optimization of the waste paper recycling process can be achieved, and the efficiency and benefit of the entire waste paper recycling industry chain can be improved. For example, according to the regional waste paper recycling quantity distribution map and the equipment reliability evaluation index, the scheduling of waste paper recycling vehicles is reasonably arranged to improve transportation efficiency; according to the maintenance condition of the equipment, the equipment replacement and repair plan in the logistics management system is optimized to reduce the impact of equipment failure on recycling work.

[0117] In terms of abnormality analysis, more types of abnormalities and analysis methods can be introduced. In addition to the analysis of periodic fluctuation patterns and sudden deviations, factors such as long-term trend changes and seasonal fluctuations of equipment can also be considered. Through analysis of these factors, more comprehensive abnormalities of the equipment can be found, and measures can be taken in advance to prevent and handle them.

[0118] For the generation of data management reports, customization can be made according to the needs of different users. For example, macro data analysis and decision-making suggestions are provided for managers, and detailed equipment failure analysis and maintenance plans are provided for technical personnel. At the same time, the visualization of the report can be improved, such as adding more charts and graphs, making the report more intuitive and easy to understand.

[0119] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system, device and unit described above can refer to the corresponding process in the foregoing method embodiments, which will not be repeated here.

[0120] It will be apparent to those skilled in the art that the application is not limited to the details of the above-exemplified embodiments and that the present application can be implemented in other particular forms without departing from the spirit or essential characteristics of the present application. The embodiments should therefore be considered in all respects as illustrative and not restrictive, the scope of the application being indicated by the appended claims rather than by the above description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Claims

1. A smart management method for waste paper weighing data based on the Internet of Things, characterized in that, The method includes: Acquire a set of real-time weighing data uploaded by multiple waste paper collection devices, wherein the set of real-time weighing data includes device identifiers and weight measurement values ​​at different time points; The real-time weighing data set is dynamically corrected to generate a standardized weighing dataset, which is then merged and the weight change trend features corresponding to each device identifier are extracted. Anomaly analysis was performed on the weight change trend characteristics to determine the type of operational anomaly of the waste paper collection equipment and the corresponding anomaly confidence level. Based on the type of operational anomaly and the anomaly confidence level, a set of equipment maintenance instructions is generated, and the set of equipment maintenance instructions is sent to the target waste paper collection equipment to trigger a calibration operation; Periodic data management reports are generated based on the calibrated standardized weighing data set.

2. The method according to claim 1, characterized in that, The step of dynamically correcting the real-time weighing data set to generate a standardized weighing dataset and extracting the weight change trend features corresponding to each device identifier includes: Identify the sequence of measured values ​​within a continuous time window in the real-time weighing data set, and calculate the difference in measured values ​​between adjacent time points; When the difference in the measured values ​​exceeds a preset fluctuation threshold, the environmental parameter set of the corresponding waste paper collection equipment is obtained. The environmental parameter set includes temperature measurement value, humidity measurement value, and equipment vibration intensity. The dynamic compensation model is invoked to assign weights to the set of environmental parameters, generating environmental disturbance correction coefficients. The measured values ​​in the real-time weighing dataset are compensated based on the environmental interference correction coefficient. After obtaining the source time points of the measured values ​​after merging and correcting the standardized weighing dataset, the weight change trend features corresponding to each device identifier are extracted.

3. The method according to claim 2, characterized in that, The step of calling the dynamic compensation model to assign weights to the set of environmental parameters and generate environmental interference correction coefficients includes: The temperature measurement value, humidity measurement value, and equipment vibration intensity are input into the pre-trained feature mapping layer to generate an environmental interference vector. The environmental disturbance vector is subjected to a multi-order nonlinear transformation to obtain the dynamic influence weight corresponding to each environmental parameter. The environmental parameter set is weighted and fused according to the dynamic influence weights to generate the environmental interference correction coefficient.

4. The method according to claim 1, characterized in that, The anomaly analysis of the weight change trend characteristics to determine the type of operational anomaly and corresponding anomaly confidence level of the waste paper collection equipment includes: Extract the periodic fluctuation patterns and sudden offsets from the weight change trend characteristics; The periodic fluctuation pattern is matched with the normal fluctuation range in historical data to generate a periodic anomaly score. The time decay coefficient is calculated for the sudden offset to generate a sudden anomaly score; The periodic anomaly score is superimposed with the sudden anomaly score to obtain a comprehensive anomaly score; The operational anomaly type and corresponding anomaly confidence level are determined based on the comprehensive anomaly score and the preset anomaly level classification rules.

5. The method according to claim 4, characterized in that, The step of calculating the time decay coefficient of the sudden offset to generate a sudden anomaly score includes: Identify the start time point and duration corresponding to the sudden offset; The time decay coefficient is determined based on the duration and a preset decay function, wherein the decay function is an exponential function that decreases as the duration increases. The sudden anomaly score is obtained by multiplying the time decay coefficient with the absolute value of the sudden offset. When the cumulative value of the sudden anomaly score exceeds the preset sudden threshold, it is marked as a device hardware failure type.

6. The method according to claim 1, characterized in that, The step of generating a set of equipment maintenance instructions based on the operational anomaly type and the anomaly confidence level includes: Based on the type of operational anomaly, a preset maintenance strategy library is matched to obtain the maintenance operation sequence and operation priority corresponding to the type of operational anomaly. The execution order of the maintenance operation sequence is dynamically adjusted based on the anomaly confidence level to generate an optimized maintenance operation sequence. The optimized maintenance operation sequence is bound to the identifier of the target waste paper collection equipment to generate the equipment maintenance instruction set; The set of equipment maintenance instructions is sorted according to operation priority and sent to the target device, and the instruction execution status is recorded.

7. The method according to claim 6, characterized in that, The step of sending the set of equipment maintenance instructions to the target device after sorting them according to operation priority and recording the instruction execution status includes: Receive instruction response data returned by the target waste paper collection equipment, wherein the instruction response data includes the maintenance operation completion time and the measurement value after the operation; The maintenance operation completion time is compared with the preset standard maintenance duration to generate a maintenance efficiency score; A maintenance effectiveness score is generated based on the deviation between the measured value after the operation and the standardized weighing data set. The maintenance efficiency score and maintenance effect score are associated and stored in the maintenance record database, and the operation priority in the maintenance strategy library is updated.

8. The method according to claim 1, characterized in that, The generation of periodic data management reports based on the calibrated standardized weighing data set includes: Statistical analysis of the frequency of abnormal occurrences and the distribution of maintenance response times for all waste paper collection equipment within a preset period; Extract the cumulative weight values ​​from the calibrated standardized weighing dataset to generate a regional waste paper recycling distribution map; Based on the frequency of anomalies and the distribution of maintenance response times, generate equipment reliability assessment indicators; The data management report containing optimization suggestions is generated by correlating the regional waste paper recycling distribution map with the equipment reliability assessment indicators. The data management report is pushed to the designated terminal and a visualization display is triggered.

9. The method according to claim 8, characterized in that, The step of extracting the cumulative weight values ​​from the standardized weighing dataset after calibration to generate a regional waste paper recycling distribution map includes: Obtain the geographic location information and area identifier of each waste paper collection device; The cumulative weight values ​​are grouped and aggregated according to the region identifier to generate the total amount of waste paper recycled in each region. The total amount of waste paper recycled is normalized with the regional population density data to generate a recycling efficiency value per unit population. A heat map is drawn based on the unit population recycling efficiency value, and areas with efficiency values ​​lower than a preset threshold are marked as areas to be optimized.

10. An intelligent management system for waste paper weighing data based on the Internet of Things, characterized in that, include: The processor, communication interface, memory, and communication bus are provided, wherein the processor, communication interface, and memory communicate with each other via the communication bus. The memory is used to store computer programs; the processor is used to execute the computer programs to implement the steps of the IoT-based intelligent management method for waste paper weighing data as described in any one of claims 1-9.

Images