A method and system for monitoring enterprise logistics and transportation

Abstract

The present invention relates to the field of monitoring and early warning, and in particular to a method and system for monitoring enterprise logistics and transportation. The method includes the following steps: calculating the noise level of each pressure data; taking any neighborhood pressure data as the cluster center, making several extension directions, and calculating the similarity of each extension direction; and calculating the similarity mutation degree according to the similarity; taking the extension direction with a similarity mutation degree greater than a preset threshold as the boundary direction, and dividing the data into several categories according to the boundary direction; calculating the classification effect, and taking the classification result corresponding to the cluster center with the largest classification effect as the final classification result; obtaining corrected data according to the noise level and the classification result; using the corrected data as the fitting expected data of the Wiener filtering algorithm to denoise the pressure data; calculating the bumpiness, and performing transportation monitoring according to the bumpiness. Accurate denoising is achieved by accurately obtaining the expected data required in the Wiener filtering algorithm.

Description

Technical Field

[0001] The present invention relates to the field of monitoring and early warning, and in particular to a method and system for monitoring enterprise logistics and transportation. Background Art

[0002] Some goods are easily damaged during transportation, such as fruits and vegetables. Especially when the road is bumpy, the damage rate is higher. In order to reduce the damage rate of goods, it is necessary to monitor the bumps during the transportation of goods, so as to remind the driver in time when the bumps exceed the tolerance of the goods.

[0003] In order to reflect the bumpy conditions of the vehicle, it is necessary to collect the pressure data in the car. However, the vehicle is subject to bumps and vibrations during driving, which will cause noise interference in the collected pressure data. In order to improve the accuracy of bump monitoring, the collected pressure data needs to be denoised.

[0004] As a filtering algorithm, the Wiener filter algorithm is often used for data denoising. Among them, the main factor affecting the accuracy of the Wiener filter algorithm is the fitting of weight parameters; and the weight parameters are generally fitted with reference to the expected signal, so the selection of the expected signal will affect the denoising accuracy of the Wiener filter algorithm. Traditionally, the signal to be denoised is often used as the expected signal, but there is noise in the signal to be denoised itself, and the accuracy of the weight parameters fitted by using it as the expected signal is low. Therefore, how to obtain an accurate expected signal becomes the research focus of this invention.

[0005] A logistics and transportation monitoring system based on the Internet of Things is disclosed in a patent application document with publication number CN112305978A. The patent application document mainly shows the hardware used and the connection relationship between the hardware, but does not reflect aspects such as denoising. Therefore, the method in the patent application document cannot effectively solve the technical problems in the present invention. Summary of the invention

[0006] In order to solve the problem of how to obtain accurate expected signals, the present invention provides an enterprise logistics transportation monitoring method and system.

[0007] In a first aspect, the present invention provides a method for monitoring enterprise logistics transportation, which adopts the following technical solution:

[0008] A method for monitoring enterprise logistics transportation, comprising the steps of:

[0009] Obtain pressure data at each location in the logistics transport vehicle at each time;

[0010] Calculating noise level , , taking the pressure data as the center, a preset number of evenly distributed extension directions are made in all directions. represents the pressure data, represents the mean value of the pressure data in the preset neighborhood of the pressure data, Indicates The similarity of the preset neighborhood pressure data in the extension direction, represents the mean value of the cosine similarity of all preset neighborhood pressure data in the two extension directions, Indicates the number of extension directions;

[0011] Taking any preset neighborhood pressure data as a feasible cluster center, taking the feasible cluster center as the center, making a preset number of uniformly distributed extension directions around it, calculating the cosine similarity of the pressure data in one extension direction and the previous extension direction as the similarity of the extension direction; calculating the similarity mutation degree; taking the extension direction with a similarity mutation degree greater than a preset threshold as the boundary direction, using the boundary direction to separate the preset neighborhood into several areas, and taking the preset neighborhood pressure data in each area as a category; calculating the classification effect, and taking the classification result corresponding to the feasible cluster center with the largest classification effect as the final classification result;

[0012] Using normalized data of the inverse of the noise level as a weight, weighted summing is performed on the pressure data in the category to which the pressure data belongs to obtain corrected data of the pressure data;

[0013] Using the corrected data as the expected fitting data, supervising the fitting process of the Wiener filter algorithm, and using the Wiener filter algorithm after the fitting to perform denoising on the pressure data;

[0014] The degree of turbulence is calculated and transport monitoring is performed by comparing the turbulence degree with a preset threshold.

[0015] The present invention takes into account that the vehicle bumps and vibrates greatly during transportation, resulting in a high possibility that the collected pressure data contains noise. Therefore, before using the pressure data to perform a bump analysis on the vehicle, a denoising process is performed; further, in the process of denoising, in order to obtain accurate expected data of each pressure data, the pressure data at the surrounding positions are used to correct the pressure data; further, when using the pressure data at the surrounding positions to correct the pressure data, considering that the pressure data at the surrounding positions may contain noise, the situation where the pressure data contains noise is analyzed, thereby determining the reference situation for each surrounding pressure data according to the situation where the pressure data at the surrounding positions contains noise; further, when analyzing the situation where each pressure data contains noise, the pressure data at the surrounding positions is corrected. In the case of noise, considering that the pressures of the same box surface at various positions on the car body should be similar, the difference between each pressure data and the pressure data at the surrounding positions is used to determine whether there is noise in each pressure data, thereby improving the accuracy of pressure noise determination; further, when analyzing the situation where each pressure data has noise, considering that the acquisition position of the pressure data may be at the boundary of the box, there may be pressure data generated by other boxes in its neighborhood pressure data, resulting in a large difference between the calculated pressure data, but this large difference is not caused by the noise in the pressure data, so when performing noise analysis, the interference of the pressure data acquisition position at the boundary of the box is excluded to improve the accuracy of noise determination. Furthermore, when using other surrounding pressure data for correction, considering that there are differences in pressure data of different boxes, it is inaccurate to use the pressure data of other boxes to correct the pressure data, and the surrounding pressure data is clustered to filter out the pressure data in the same box as the pressure data. Furthermore, when clustering the surrounding pressure data, considering that the edges of the boxes are straight lines, the pressure data of different boxes can be separated by the dividing lines. Therefore, classification processing is performed by obtaining the dividing lines. This classification method has higher classification accuracy and higher implementation efficiency.

[0016] Preferably, the calculating of the degree of turbulence comprises:

[0017] The pressure data of a position at all times are processed in segments to obtain a number of data segments of the position;

[0018] Obtain the pressure data of each position in the logistics transport vehicle after loading is completed but before driving, and record it as the baseline pressure data;

[0019] Calculate the degree of turbulence based on the segmented results:

[0020]

[0021] in, Indicates The number of data segments at each location, represents the number of locations where pressure data is collected, Indicates The first position The maximum value of the denoised pressure data of the data segment, Indicates Baseline pressure data for each location, Indicates collection of The first position The length of time required for the data in a data segment, Indicates The number of data segments at each location, Indicates the degree of turbulence.

[0022] The present invention accurately reflects the damage of cargo caused by bumps through the change of pressure data within a unit time, thereby providing a benchmark for accurate bump monitoring.

[0023] Preferably, the step of segmenting the pressure data of a position at all times includes:

[0024] The sequence consisting of the denoised pressure data at all times of a position is recorded as the analysis sequence;

[0025] The data that are continuously smaller than the reference pressure data in the analysis sequence are segmented together, and the data that are continuously not smaller than the reference pressure data are segmented together to obtain a number of data segments.

[0026] The present invention divides the pressure data collected under the same movement trend of the cargo into an interval, thereby providing a basis for accurately analyzing the damage to the cargo caused by bumps.

[0027] Preferably, the method for obtaining the preset neighborhood pressure data includes:

[0028] Get the length and width of each box side in the logistics transport vehicle as the reference length and base width ;

[0029] Taking the collection position of the pressure data as the center, The area is taken as a preset neighborhood, and the pressure data of each collection position in the preset neighborhood is recorded as neighborhood pressure data of the pressure data.

[0030] The present invention sets the neighborhood through the side of the box, so that the neighborhood data collected in this way is more likely to be the pressure data generated by the same box, providing a basis for accurately analyzing the noise situation.

[0031] Preferably, the method for obtaining the similarity includes:

[0032] The inverse of the data variance is taken as the similarity.

[0033] Preferably, the method for obtaining the similarity mutation degree includes:

[0034] The difference between the similarity of one extension direction and the similarity of the previous extension direction is taken as the similarity mutation degree.

[0035] Preferably, the calculating classification effect includes:

[0036] The mean value of the cosine similarity of all preset neighborhood pressure data in two extension directions of the same category is recorded as the similarity of the same category;

[0037] The average of the cosine similarities of the preset neighborhood pressure data in two extension directions of different categories is taken as the heterogeneous similarity;

[0038] The quotient of the similarity of the same category and the similarity of different categories is taken as the classification effect.

[0039] Preferably, the transport monitoring by comparing the degree of turbulence with a preset threshold value includes:

[0040] comparing the degree of bumpiness with a preset threshold;

[0041] If the degree of turbulence is greater than the preset threshold, the driver will be reminded to slow down;

[0042] If the degree of turbulence is not greater than the preset threshold, no reminder will be given.

[0043] In a second aspect, the present invention provides an enterprise logistics transportation monitoring system, which adopts the following technical solution:

[0044] A system for monitoring enterprise logistics and transportation includes: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned method for monitoring enterprise logistics and transportation is implemented.

[0045] By adopting the above technical solution, the above-mentioned enterprise logistics transportation monitoring method is generated into a computer program and stored in a memory to be loaded and executed by a processor, so that a terminal device is made according to the memory and the processor for easy use.

[0046] The present invention has the following technical effects:

[0047] The present invention takes into account that the vehicle is subject to large bumps and vibrations during transportation, which may lead to a high possibility of noise in the collected pressure data. Therefore, before using the pressure data to analyze the bumps of the vehicle, denoising is performed;

[0048] Furthermore, in the process of denoising, in order to obtain accurate expected data of each pressure data, the pressure data at the surrounding positions are used to correct the pressure data;

[0049] Furthermore, when the pressure data at the surrounding positions are used to correct the pressure data, considering that the pressure data at the surrounding positions may have noise, the situation where the pressure data has noise is analyzed, thereby determining the reference situation for each surrounding pressure data according to the situation where the pressure data at the surrounding positions has noise;

[0050] Furthermore, when analyzing the situation where each pressure data has noise, considering that the pressures of each position on the same box surface on the car should be similar, the situation where each pressure data has noise is judged based on the difference between each pressure data and the pressure data of the surrounding positions, thereby improving the accuracy of the pressure noise judgment;

[0051] Furthermore, when analyzing the noise situation of each pressure data, taking into account that the collection position of the pressure data may be at the boundary of a box, there may be pressure data generated by other boxes in its neighborhood pressure data, which leads to large differences between the calculated pressure data. However, this large difference is not caused by the noise in the pressure data. Therefore, when performing noise analysis, the interference of the pressure data collection position at the boundary of the box is excluded to improve the accuracy of noise judgment.

[0052] Furthermore, when using other surrounding pressure data for correction, considering the differences in pressure data in different boxes, it is inaccurate to correct the pressure data using the pressure data of other boxes, so the surrounding pressure data are clustered to filter out the pressure data in the same box as the pressure data.

[0053] Furthermore, when clustering the surrounding pressure data, considering that the edges of the boxes are straight lines, the pressure data of different boxes can be separated by the dividing lines. Therefore, classification processing is performed by obtaining the dividing lines. This classification method has higher classification accuracy and higher implementation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0055] Figure 1 It is a flow chart of a method in a method for monitoring enterprise logistics transportation according to an embodiment of the present invention;

[0056] Figure 2 A schematic diagram of the relationship between the pressure data collection location and box distribution provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0058] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0059] The embodiment of the present invention discloses a method for monitoring enterprise logistics transportation, referring to Figure 1 , comprising steps S1 to S4:

[0060] S1: Obtain the pressure data of each location in the logistics transport vehicle at each time.

[0061] Specifically, pressure sensors are installed at several locations in the logistics transport vehicle, and the pressure data at each moment is collected by using the pressure sensors.

[0062] S2: Calculate the noise level; classify the pressure data and preset neighborhood pressure data, use the normalized data of the inverse of the noise level as a weight, and perform weighted summation on the pressure data in the category where the pressure data belongs to obtain the corrected data of the pressure data.

[0063] It should be noted that the expected data required for fitting the weight parameters in the Wiener filter algorithm is generally useful data with less noise. In order to obtain the expected data, relatively accurate pressure data must be obtained. In order to resist collision, the goods are generally placed in a box for transportation. Therefore, the box, as a basic transportation unit, can be regarded as a whole, and the pressure of each position on the logistics transport vehicle is similar. Therefore, based on this feature, the pressure data of the surrounding positions can be used to correct the pressure data to obtain relatively accurate pressure data.

[0064] It should be further explained that, since the pressure data at the surrounding positions may also contain noise, when making corrections with reference to the pressure data at the surrounding positions, the fact that the pressure data at the surrounding positions contain noise must be considered.

[0065] S20: Calculate the noise level.

[0066] It should be noted that when there is no noise in the pressure data, the similarity between the pressure data and the surrounding pressure data should be large, so noise determination can be performed based on this. Figure 2 The relationship between the pressure data collection location and the box distribution is shown in the diagram. If the pressure data collection location is at the border of the box, there will be some pressure data around the pressure data that are not in the same box as the projection area. At this time, the calculated pressure data may be quite different. This is not caused by noise, so when performing noise analysis, the interference of this situation needs to be eliminated.

[0067] Preferably, as an example, calculating the noise level includes:

[0068]

[0069] Wherein, taking the pressure data as the center, a preset number of evenly distributed extension directions are made in all directions. represents the pressure data, represents the mean value of the pressure data in the preset neighborhood of the pressure data, Indicates The similarity of the preset neighborhood pressure data in the extension direction, represents the mean value of the cosine similarity of all preset neighborhood pressure data in the two extension directions, Indicates the number of extension directions, Indicates the noise level of the pressure data.

[0070] Understandably, The larger the value is, the greater the difference between the pressure data and the surrounding pressure data is. Under normal circumstances, the difference between the pressure data and the surrounding pressure data should be small. Therefore, when the difference between the pressure data and the surrounding pressure data is large, the pressure data is more affected by noise. When the collection position of the pressure data is at the boundary of the projection areas of different boxes, even if there is no noise in the pressure data, there will be a large difference between the pressure data and the surrounding pressure data. When the pressure data is at the boundary of the projection areas of different boxes, the pressure data at the collection positions within the projection areas of the same box should be similar, and the difference between the pressure data at the collection positions within the projection areas of different boxes should be large. Figure 2 As shown, the projection area of ​​a box will be concentrated. If the pressure data is on the boundary of the projection area of ​​different boxes, the pressure data in the extension direction starting from the pressure data is likely to be distributed inside the projection area of ​​the same box, so the similarity of the pressure data in the extension direction starting from the pressure data should be greater. Figure 2The pressure data shown is on the boundary of different boxes, and there will be projection areas of different boxes around the pressure data. Therefore, different extension directions are likely to be distributed in different projection areas. Therefore, the difference of pressure data in different extension directions should be large. Therefore, the situation that the pressure data is on the boundary of the projection area of ​​different boxes can be judged based on the similarity of data in the same extension direction and the difference of data in different extension directions. This reflects the situation where the pressure data is at the projection boundary of different boxes. Divide by To eliminate the interference of the pressure data noise situation determination when the pressure data is at the projection boundary of different boxes.

[0071] It should be noted that the number of collection positions passed through the two extension directions may be different. When calculating the cosine similarity, it is only necessary to obtain the number of collection positions in the extension direction with fewer collection positions in the two extension directions as the comparison number, and randomly extract the pressure data of the comparison number of collection positions from all the collection positions in the other extension direction of the two extension directions, and use the pressure data extracted in the two extension directions to calculate the cosine similarity.

[0072] The above embodiments involve preset neighborhood pressure data, and a method for determining the preset neighborhood pressure data will be described below.

[0073] It should be noted that, since the pressure data within the projection area of ​​a box have relatively large similarity in values, the pressure data within the projection area of ​​a box is used as the preset neighborhood pressure data.

[0074] The method for obtaining the preset neighborhood pressure data includes:

[0075] Get the length and width of each box side in the logistics transport vehicle as the reference length and base width ;

[0076] Taking the collection position of the pressure data as the center, The area is taken as a preset neighborhood, and the pressure data of each collection position in the preset neighborhood is recorded as neighborhood pressure data of the pressure data.

[0077] It can be understood that the side of each box in the logistics transport vehicle is the projection area of ​​each box on the side of the carriage of the logistics transport vehicle.

[0078] S21: Classify the pressure data and preset neighborhood pressure data.

[0079] It should be noted that since the pressure data of the collection positions under the projection area of ​​the same box are more similar, the accuracy of correction by referring to the pressure data of the collection positions under the projection area of ​​the same box is higher, so the pressure data needs to be clustered.

[0080] Optionally, as an example, the pressure data and the preset neighborhood pressure data are classified, including:

[0081] The pressure data and the preset neighborhood pressure data are clustered using the DBSCAN clustering algorithm to obtain several categories.

[0082] It should be noted that the DBSCAN clustering algorithm has a parameter adjustment process and the clustering efficiency of the algorithm is low.

[0083] It should be further explained that the projection area of ​​each box is a rectangle, so the boundary between different categories is a straight line, so different categories can be separated by a straight line. To determine a straight line, you only need to determine a point it passes through and the direction of the straight line. Therefore, the following is classified according to this principle.

[0084] Preferably, as an example, the pressure data and the preset neighborhood pressure data are classified, including:

[0085] Taking any preset neighborhood pressure data as a feasible cluster center, taking the feasible cluster center as the center, making a preset number of uniformly distributed extension directions around it, calculating the cosine similarity of the pressure data in one extension direction and the previous extension direction as the similarity of the extension direction; calculating the similarity mutation degree;

[0086] The extension direction in which the similarity mutation degree is greater than the preset threshold is used as the demarcation direction, and the preset neighborhood is divided into several areas using the demarcation direction, and the preset neighborhood pressure data in each area is taken as a category; in this embodiment, the mean of the similarities of all extension directions is used as the preset threshold, and other embodiments may set other preset thresholds, which are not specifically limited in this embodiment.

[0087] Calculate the classification effect, and take the classification result corresponding to the feasible cluster center with the largest classification effect as the final classification result.

[0088] It can be understood that the pressure data of the projection areas of different boxes are quite different, and the pressure data of the projection areas of the same box are quite similar. Therefore, compared with the two extension directions being in the projection area of ​​the same box, when the two extension directions are in the projection areas of two boxes, their similarity suddenly decreases, so the direction of the straight line can be obtained by the degree of similarity mutation. At the same time, in order to determine the passing point of the straight line, it is necessary to judge its classification effect, and determine the accurate passing point of the straight line by the classification effect.

[0089] The above embodiments involve the similarity mutation degree and the classification effect. The following describes the method for determining the similarity mutation degree and the classification effect.

[0090] The method for obtaining the similarity mutation degree includes: taking the difference between the similarity in one extension direction and the similarity in the previous extension direction as the similarity mutation degree.

[0091] In addition, the method for obtaining the classification effect includes: recording the mean value of the cosine similarity of all preset neighborhood pressure data in two extension directions of the same category as the similarity of the same category;

[0092] The average of the cosine similarities of the preset neighborhood pressure data in two extension directions of different categories is taken as the heterogeneous similarity;

[0093] The quotient of the similarity of the same category and the similarity of different categories is taken as the classification effect.

[0094] S22: using normalized data of the inverse of the noise level as a weight, and performing weighted summation on the pressure data in the category to which the pressure data belongs to obtain corrected data of the pressure data.

[0095] It should be noted that pressure data with a smaller noise level has a higher accuracy, and the pressure data should be corrected with reference to the surrounding pressure data.

[0096] This embodiment uses a linear normalization method to normalize the inverse of the noise level. Other embodiments may use other normalization methods, and this embodiment does not specifically limit this.

[0097] S3: Using the corrected data as the expected fitting data, supervising the fitting process of the Wiener filter algorithm, and using the Wiener filter algorithm after the fitting to perform denoising on the pressure data.

[0098] Preferably, as an example, the corrected data is used as the expected fitting data to supervise the fitting process of the Wiener filter algorithm, and the pressure data is denoised using the Wiener filter algorithm after the fitting, including:

[0099] The corrected data is used as the expected fitting data of the pressure data, the parameters in the Wiener filter algorithm are fitted with reference to the expected fitting data, and the pressure data is denoised based on the fitted Wiener filter algorithm.

[0100] It can be understood that by providing a relatively accurate expected data, the weight parameters fitted by the Wiener filter algorithm can be more accurate, thereby improving the accuracy of denoising.

[0101] S4: Calculate the degree of turbulence and monitor the transportation by comparing the degree of turbulence with a preset threshold.

[0102] S40: Calculate the degree of turbulence.

[0103] It should be noted that when a vehicle is bumpy, the vehicle will move up and down and left and right. Therefore, the bumping of the vehicle can be analyzed based on the up and down and left and right movements of the vehicle.

[0104] Preferably, as an example, calculating the degree of turbulence includes:

[0105] Obtain the pressure data of each position in the logistics transport vehicle after loading is completed but before driving, and record it as the baseline pressure data;

[0106] The sequence consisting of the denoised pressure data at all times of a position is recorded as the analysis sequence;

[0107] The data that are continuously smaller than the reference pressure data in the analysis sequence are segmented together, and the data that are continuously not smaller than the reference pressure data are segmented together to obtain a number of data segments.

[0108] The degree of turbulence satisfies the relationship:

[0109]

[0110] in, Indicates The number of data segments at each location, represents the number of locations where pressure data is collected, Indicates The first position The maximum value of the denoised pressure data of the data segment, Indicates Baseline pressure data for each location, Indicates collection of The first position The length of time required for the data in a data segment, Indicates The number of data segments at each location, Indicates the degree of turbulence.

[0111] It is understandable that the baseline pressure data reflects the pressure data when the vehicle is not bumpy. If the pressure data is or is less than the baseline pressure data, it means that the vehicle is bumpy. Therefore, the pressure data of the cargo under the same movement trend can be segmented by using the baseline pressure data. For example, if the pressure data is continuously greater than the baseline pressure data, it means that the bumpy vehicle causes the cargo to move toward the wall or the bottom of the car, but due to the obstruction of the car wall, it cannot move and can only generate elastic potential energy. When the cargo undergoes elastic deformation, it will cause damage. Therefore, the bumpy condition of the vehicle can be measured by analyzing the loss of the cargo caused by the bump. It reflects the impact of elastic force on the goods. The larger the value is, the greater the impact of elastic force on the goods is and the greater the damage to the goods is. It reflects the number of times the goods undergo elastic deformation.

[0112] S41: Transport monitoring is performed by comparing the degree of turbulence with a preset threshold.

[0113] Preferably, as an example, transport monitoring is performed by comparing the degree of turbulence with a preset threshold, including:

[0114] comparing the degree of bumpiness with a preset threshold;

[0115] If the degree of turbulence is greater than the preset threshold, the driver will be reminded to slow down;

[0116] If the degree of turbulence is not greater than the preset threshold, no reminder will be given.

[0117] An embodiment of the present invention further discloses an enterprise logistics and transportation monitoring system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an enterprise logistics and transportation monitoring method according to the present invention is implemented.

[0118] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0119] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory, a dynamic random access memory, a static random access memory, an enhanced dynamic random access memory, a high bandwidth memory, a hybrid storage cube, etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of the device or accessible or connectable to the device.

[0120] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0121] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for monitoring enterprise logistics and transportation, characterized in that: Includes steps: Obtain pressure data at each location in the logistics transport vehicle at each time; Calculate the noise level of pressure data , , taking the pressure data as the center, a preset number of evenly distributed extension directions are made in all directions. represents the pressure data, represents the mean value of the pressure data in the preset neighborhood of the pressure data, Indicates The similarity of the preset neighborhood pressure data in the extension direction, represents the mean value of the cosine similarity of all preset neighborhood pressure data in the two extension directions, Indicates the number of extension directions; Taking any preset neighborhood pressure data as a feasible cluster center, taking the feasible cluster center as the center, making a preset number of uniformly distributed extension directions around it, calculating the cosine similarity of the pressure data in one extension direction and the previous extension direction as the similarity of the extension direction; calculating the similarity mutation degree; The extending direction where the similarity mutation degree is greater than the preset threshold is used as the dividing direction, the preset neighborhood is divided into several regions using the dividing direction, and the preset neighborhood pressure data in each region is used as a category; Calculate the classification effect, and take the classification result corresponding to the feasible cluster center with the largest classification effect as the final classification result; Using normalized data of the inverse of the noise level as a weight, weighted summing is performed on the pressure data in the category to which the pressure data belongs to obtain corrected data of the pressure data; Using the corrected data as the expected fitting data, supervising the fitting process of the Wiener filter algorithm, and using the Wiener filter algorithm after the fitting to perform denoising on the pressure data; Calculating the degree of bumpiness includes: segmenting the pressure data of a position at all times to obtain several data segments of the position; obtaining the pressure data of each position in the logistics transport vehicle after loading is completed but before driving as the reference pressure data; and calculating the degree of bumpiness based on the segmentation results: in, Indicates The number of data segments at each location, represents the number of locations where pressure data is collected, Indicates The first position The maximum value of the denoised pressure data of the data segment, Indicates Baseline pressure data for each location, Indicates collection of The first position The length of time required for the data in a data segment, Indicates The number of data segments at each location, Indicates the degree of turbulence, and transport monitoring is performed by comparing the turbulence degree with a preset threshold.

2. The enterprise logistics and transportation monitoring method according to claim 1 is characterized in that: The step of segmenting the pressure data of a position at all times includes: The sequence consisting of the denoised pressure data at all times of a position is recorded as the analysis sequence; The data that are continuously smaller than the reference pressure data in the analysis sequence are segmented together, and the data that are continuously not smaller than the reference pressure data are segmented together to obtain a number of data segments.

3. The enterprise logistics and transportation monitoring method according to claim 1 is characterized in that: The method for obtaining the preset neighborhood pressure data includes: Get the length and width of each box side in the logistics transport vehicle as the reference length and base width ; Taking the collection position of the pressure data as the center, The area is taken as a preset neighborhood, and the pressure data of each collection position in the preset neighborhood is recorded as neighborhood pressure data of the pressure data.

4. The enterprise logistics transportation monitoring method according to claim 1 is characterized in that: The method for obtaining the similarity includes: The inverse of the data variance is taken as the similarity.

5. The enterprise logistics and transportation monitoring method according to claim 1 is characterized in that: The method for obtaining the similarity mutation degree includes: The difference between the similarity of one extension direction and the similarity of the previous extension direction is taken as the similarity mutation degree.

6. The enterprise logistics and transportation monitoring method according to claim 1 is characterized in that: The calculating classification effect includes: The mean value of the cosine similarity of all preset neighborhood pressure data in two extension directions of the same category is recorded as the similarity of the same category; The average of the cosine similarities of the preset neighborhood pressure data in two extension directions of different categories is taken as the heterogeneous similarity; The quotient of the similarity of the same category and the similarity of different categories is taken as the classification effect.

7. The enterprise logistics and transportation monitoring method according to claim 1 is characterized in that: The transport monitoring by comparing the degree of turbulence with a preset threshold value includes: comparing the degree of bumpiness with a preset threshold; If the degree of turbulence is greater than the preset threshold, the driver will be reminded to slow down; If the degree of turbulence is not greater than the preset threshold, no reminder will be given.

8. An enterprise logistics and transportation monitoring system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, an enterprise logistics and transportation monitoring method according to any one of claims 1-7 is implemented.

Images