An intelligent disassembly platform for waste motor vehicles

Abstract

The present invention relates to the technical field of motor vehicle disassembly. Specifically, it relates to an intelligent disassembly platform for waste motor vehicles, including a motor vehicle transportation module, an artificial intelligence acquisition module, a robot disassembly module, an automated storage module, and a user interaction interface module. The present invention collects the image features of waste motor vehicle parts through an image acquisition unit, determines the position and part signal of the motor vehicle parts according to the image features. The robot disassembly module adjusts the automobile disassembly data based on the above data. When the image acquisition unit cannot collect image data, the weight acquisition unit judges the parts according to the weight of the motor vehicle parts. After the weight determines the part type, the part shape is updated to the image acquisition unit. At the same time, when the weight acquisition unit collects that the part weight does not conform to the part threshold after the image acquisition unit collects the part picture, it indicates that the internal components of the part are missing. At this time, the robot disassembly module is adjusted according to the part loss situation to control the grasping force of the part.

Description

Technical Field

[0001] The present invention relates to the technical field of motor vehicle disassembly, and more specifically, to an intelligent disassembly platform for waste motor vehicles. Background Art

[0002] A scrapped motor vehicle is a motor vehicle that reaches the national scrapping standard or, although it has not reached the national scrapping standard, has a severely damaged engine or chassis and fails to meet the national motor vehicle operation safety technical conditions or the national motor vehicle pollutant emission standards after inspection, and is called a scrapped motor vehicle;

[0003] Existing technologies usually rely on an image acquisition module to identify and classify motor vehicle parts. In this process, the machine vision system analyzes the images to determine the types of parts, and mechanical grippers are used for sorting. However, during this operation, the components inside the parts are prone to be lost or damaged due to improper grasping force of the mechanical grippers, especially for parts with complex structures or fragile materials;

[0004] After a waste motor vehicle has been stored for a long time, rust often appears on its surface. Existing technologies mainly focus on removing the rust on the surface of the parts. However, rust is also a reusable resource of waste motor vehicles. However, when disassembling waste motor vehicles, a small amount of rust will fall off from the motor vehicle parts, affecting the judgment of the recycling value of the motor vehicle parts. Summary of the Invention

[0005] The purpose of the present invention is to provide an intelligent disassembly platform for waste motor vehicles to solve the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides an intelligent disassembly platform for waste motor vehicles, including a motor vehicle transportation module, an artificial intelligence acquisition module, a robot disassembly module, an automated storage module, and a user interface module;

[0007] The motor vehicle transportation module is used to transport the motor vehicle through an automated guided vehicle;

[0008] The automated guided vehicle automatically travels according to a preset route, automatically guides and scans the surrounding environment through lidar to construct a three-dimensional map, detects obstacles, identifies ground markings or specific objects through a camera, measures short distances with ultrasonic sensors to detect obstacles in front, matches the sensor data with a pre-established map to determine its own position, plans an optimal path based on the current position and the destination, and controls the steering and speed of the AGV according to the instructions provided by the navigation system.

[0009] The artificial intelligence acquisition module includes an image acquisition unit, a weight acquisition unit, a data update unit, and a part judgment unit;

[0010] The image acquisition unit is used to identify motor vehicle parts and their positions through multiple cameras, and output the part position signals to the robot disassembly module;

[0011] The method for identifying motor vehicle parts is as follows:

[0012] After the camera identifies the motor vehicle parts, compare the information of the motor vehicle parts identified by the camera with the part information in the part database. After the similarity is greater than 80%, the parts are determined, where the database is referenced by the part data of the motor vehicle manufacturer;

[0013] First, from the feature vector of the part image to be identified and the reference part image in the part database , the feature vector contains descriptive information such as color, texture, and shape;

[0014] Calculate the distance D between the two feature vectors. The calculation formula is as follows:

[0015] ;

[0016] Convert the feature distance to similarity. The similarity calculation formula is:

[0017] ;

[0018] is the predefined maximum possible distance, set as the sum of the maximum values of each dimension of the feature vector;

[0019] Compare the calculated similarity S with the preset threshold T. The preset threshold T is 80%. The comparison method is as follows:

[0020] If it is determined to be similar parts if S>T, it is determined to be similar parts;

[0021] It is determined to be similar parts;

[0022] The method for identifying the position of the part is as follows:

[0023] Suppose there are two cameras, and their internal and external parameters are known. The internal parameters include the focal length ( , ) and the principal point coordinates ( , ), and the external parameters include the rotation matrix (R) and the translation vector (T);

[0024] For two cameras, if they capture images of the same part, then the projection points of the part on the image planes of the two cameras are respectively ( , ) and ( , ), the three - dimensional coordinates of the part can be calculated through the following steps;

[0025] Calculate the projection matrix: The projection matrix P of each camera can be obtained by calculating the internal parameters and external parameters:

[0026] ;

[0027] where K is the internal parameter matrix, and [R][T] is the external parameter matrix;

[0028] Establish a system of equations: According to the projection matrix, establish the following system of equations to solve for the three - dimensional coordinates:

[0029] ;

[0030] By solving this system of equations, the three - dimensional coordinates (X, Y, Z) of the part can be obtained;

[0031] Convert the camera coordinate system to the actual world coordinate system, and the calculation formula is as follows:

[0032] ;

[0033] where, ( , , ) is the actual world coordinate, ( , , ) is the coordinate in the camera coordinate system, is the rotation matrix, is the translation vector;

[0034] The weight acquisition unit is used to measure the size of the motor vehicle part through laser ranging technology, and estimate the weight of the part by combining data such as material density;

[0035] After the image acquisition unit acquires the information of the motor vehicle part, the weight acquisition unit weighs the part. If the part weight does not correspond to the part threshold, it means that the internal components of the part are missing, and the part loss signal is output to the robot disassembly module;

[0036] Use a laser rangefinder to accurately measure the key dimensions of the motor vehicle part. The laser rangefinder determines the distance by emitting a laser pulse and measuring the time it takes to reflect back. Calculate the geometric shape of the part based on the measured dimension data. According to the geometric shape and dimension data of the part, use the corresponding geometric formula to calculate the volume of the part. Query the density of the material used for the part from the database, and calculate the weight of the part through the formula;

[0037] Volume calculation formula:

[0038] For regular shapes, the formula for calculating the volume V is as follows:

[0039] ;

[0040] ;

[0041] ;

[0042] Weight estimation formula:

[0043] ;

[0044] W is the weight of the part, V is the volume of the part, and p is the density of the material;

[0045] Example

[0046] Suppose we measure the dimensions of a motor vehicle part using laser ranging technology as follows:

[0047] ;

[0048] ;

[0049] ;

[0050] ;

[0051] Calculate the volume:

[0052]

[0053] Calculate the weight:

[0054]

[0055] Therefore, the weight of the motor vehicle part is estimated to be kilograms;

[0056] The method for comparing the part weight with the part threshold is as follows:

[0057] Weight difference calculation:

[0058]

[0059] Among them, is the weight difference, is the actually measured part weight, is the standard weight threshold preset through the manufacturer's part data;

[0060] If is greater than or equal to a preset error threshold , it is considered that internal components of the part are missing;

[0061]

[0062] Assume the weight of a standard part is kilograms, and the weight difference threshold set by the system is kilograms; if during the weighing process, the actually measured weight is kilograms, then the weight difference is kilograms, which exceeds the threshold , and the system will determine that internal components of the part are missing;

[0063] The data update unit is configured to, when the image acquisition unit fails to acquire part information, collect and compare the weight of the motor vehicle part through the weight acquisition unit, determine the part weight information after the comparison, respectively distinguish the part shape information according to the part weight information, and update the part shape signal at this time to the database in the image acquisition unit after the distinction is completed;

[0064] According to the weight information, the following formula may be used to infer the part shape information: ,

[0065] where f is a function for inferring shape based on weight, and the technical formula of this function is as follows:

[0066]

[0067] where, , , …, regression coefficients, is the error term

[0068] Working principle:

[0069] When a new part weight is measured, we use the above formula to predict its shape characteristics ;

[0070] By solving the equation = to infer the shape characteristics;

[0071] The shape characteristics can then be used to determine the shape of the part;

[0072] If the inference is successful, the part information in the database is updated: number;

[0073] The part judgment unit is configured to judge whether a motor vehicle part is missing according to the image acquisition unit, and output a part missing signal to the robot disassembly module;

[0074] The absence of parts indicates that both the image acquisition unit and the weight acquisition unit have detected part information;

[0075] Working principle: The image acquisition unit obtains image information by photographing relevant parts of a motor vehicle, analyzes and processes the images to determine the presence of parts. The weight acquisition unit measures the weight of relevant parts of the motor vehicle, and the part judgment unit synthesizes the information provided by these two units to determine whether parts are missing;

[0076] If parts are missing, it means that the image acquisition unit has not detected the features of specific parts in the images it has acquired, and at the same time, there is a difference between the weight measured by the weight acquisition unit and the weight that this part should have in the complete state;

[0077] Calculation formula: Assume that the standard weight of a certain part of a complete motor vehicle is W standard, the number of parts that the image acquisition unit can identify is N image, the actual weight measured by the weight acquisition unit is W actual, and the average weight of each part is known as W average;

[0078] Then it is possible to judge whether parts are missing through the following formula: If N image is less than the expected number of parts,

[0079] and , then it is determined that parts are missing.

[0080] The robot disassembly module includes a part grasping unit, a stable signal input unit, and a running trajectory modification unit;

[0081] The part grasping unit is used to classify and disassemble motor vehicle parts through an intelligent gripper according to the shape, color, and size of the parts identified by the artificial intelligence acquisition module;

[0082] Distance calculation: ;

[0083] Angle calculation: ;

[0084] where and are the starting point and ending point coordinates respectively;

[0085] Grasping force calculation:

[0086]

[0087] where F is the grasping force, m is the part mass, and g is the acceleration due to gravity;

[0088] The stable signal input unit is used to stably receive the part position signal, part loss signal, and part absence signal output from the artificial intelligence acquisition module through radio frequency;

[0089] The radio frequency alarm signal generates a radio wave signal containing alarm information through the transmitter in the alarm system. This signal is generated by an oscillator and transmitted through an antenna; the radio wave propagates in the air and penetrates walls and other obstacles; the receiving antenna of the power output module captures the radio wave and transfers it to the internal receiving circuit, and the receiving circuit decodes the received radio wave signal to identify the alarm information;

[0090] The operating trajectory modification unit is used to modify the operating trajectory of the part grasping unit according to the signal of the stable signal input unit, where the operating trajectory includes the grasping position and clamping force of the mechanical gripper;

[0091] Examples of calculation formulas are as follows:

[0092] Assume the initial grasping position is , according to the signal of the stable signal input unit, the fine-tuning calculation formula is as follows:

[0093] , ,

[0094] The adjusted signal is:

[0095] where, , , is the signal input by the stable signal input unit;

[0096] Adjustment of clamping force:

[0097] Assume the part weight changes from to , and the corresponding clamping force changes from to , then the clamping force change value Formula calculation:

[0098] ;

[0099] where:

[0100] K is the proportionality coefficient of clamping force and weight,

[0101] is the part weight after change;

[0102] is the reference weight;

[0103] Example:

[0104] Assume the reference weight , the corresponding reference clamping force , proportionality coefficient ,

[0105] If the weight of the part increases to , then the change value of the clamping force is:

[0106]

[0107] Therefore, the clamping force needs to be increased by to adapt to the increase in the weight of the part.

[0108] The robot disassembly module further includes an air purification unit for treating harmful substances generated during the classification and disassembly process of the part grasping unit by air filtration method;

[0109] The working principle of the air filtration method is to filter and adsorb harmful substances in the air through a filter, so as to achieve the purpose of purifying the air:

[0110] The effect of the air filtration method is estimated by the following formula:

[0111] ;

[0112] Among them, the filtration efficiency represents the ability of the filter to remove harmful substances.

[0113] The automated storage module includes a classification unit, a part storage unit and a loss unit;

[0114] The classification unit is used to classify the parts disassembled by the robot disassembly module, and the classification threshold is calculated based on the combination of the shape and weight of the parts in the artificial intelligence acquisition module

[0115] Analyze the average values of the part shapes and weights in the database of the artificial intelligence acquisition module;

[0116] Comprehensively evaluate the part classification threshold according to the average estimated values of weight and shape, and because the part shape and weight data in the database are updated in real time, the part classification threshold changes in real time according to the changes in the database;

[0117] Classification threshold calculation formula:

[0118] Suppose we have parts, each part has a shape feature and a weight feature ;

[0119] Calculate the average values of shape and weight:

[0120] ;

[0121] ;

[0122] Suppose we use the linear weighted method to calculate the comprehensive score , where and are the weights of shape and weight, both being 0.5:

[0123]

[0124] is the classification threshold;

[0125] The part storage unit is used to store the parts classified by the classification unit, and the same parts are divided into different qualities;

[0126] Set the classification threshold according to the comprehensive score of the classification unit, and the quality differentiation method is as follows: Set the classification threshold, and the quality differentiation method is as follows:

[0127] is excellent;

[0128] is poor;

[0129] Among them, and are the comprehensive scores of the parts;

[0130] The loss unit is used to compare the weight of the same category of parts stored in the part storage unit with the weight detected by the artificial intelligence acquisition module, judge the change in weight during the part disassembly process, speculate the weight of the rust on the part surface, and then output the rust change signal to the user interaction interface module;

[0131] The weight calculation formula is as follows:

[0132] ;

[0133] is the total weight of the same category of parts in the storage unit, is the weight detected by the artificial intelligence acquisition module, and by analyzing the change trend of, speculate the rust situation.

[0134] The automated storage module further includes a logistics transportation unit, and the logistics transportation unit is used to automatically notify the logistics platform for cleaning when the part storage unit is full.

[0135] The user interaction interface module includes a data analysis unit and a recycling value comparison unit;

[0136] The data analysis unit is used to record and analyze the data of part shape change, weight change, and weight change caused by rust on the part surface, and determine the recycling value threshold of motor vehicle parts according to the analyzed data;

[0137] Assume the original value of the part is , then the recycling value threshold V is calculated by the following formula:

[0138]

[0139] a is the shape change coefficient, is the weight change coefficient, is the rust weight change coefficient;

[0140] The recycling value comparison unit is used to compare the recycling value and judge the recycling value threshold of motor vehicle parts according to the data analysis unit;

[0141] The recycling value is of the recycling value threshold, and the classification unit calculates of the classification threshold;

[0142] If , then the recycling value of this part is high;

[0143] If , then the recycling value of this part is low.

[0144] The user interface module also includes a display unit for sorting the recycling values of motor vehicle parts in the descending order in the recycling value comparison unit and remotely outputting them to the staff through radio frequency.

[0145] The image acquisition unit acquires the image features of waste motor vehicle parts, judges the position and signals of motor vehicle parts according to the image features. When the image acquisition unit fails to identify accurately, the weight acquisition unit judges the part type according to the weight of the motor vehicle part and acquires the part shape to update the data of the image acquisition unit;

[0146] When the part threshold collected by the weight acquisition unit does not correspond to the part weight, if the image acquisition unit acquires the information of the motor vehicle part, it means that the internal components of the part are missing;

[0147] When the image acquisition unit cannot distinguish the part by shape and weight, it means that the part is missing;

[0148] The stable signal input unit modifies the running trajectory of the part grasping unit through the part position signal, part loss signal, and part missing signal;

[0149] Compare the weight of the parts in the part storage unit with the weight of the artificial intelligence acquisition module to judge the change in the weight of the parts. The recycling value comparison unit judges the recycling value of the parts, and the threshold value of the recycling value is judged by modifying the data through the robot disassembly module and the weight change data in the loss unit.

[0150] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0151] 1. In the intelligent disassembly platform for waste motor vehicles, the image acquisition unit acquires the image features of the parts of the waste motor vehicle, judges the position and part signals of the motor vehicle parts according to the image features. The robot disassembly module modifies the disassembly data of the vehicle based on the above data. And when the weight acquisition unit cannot acquire image data by the image acquisition unit, it judges the parts according to the weight of the motor vehicle parts. After determining the part type by the weight at this time, it updates the part shape to the image acquisition unit, which is convenient for the image acquisition unit to identify the parts. At the same time, when the weight acquisition unit acquires the part picture and the weight of the part does not meet the part threshold, it indicates that the internal components of the part are lost. At this time, adjust the grasping force of the robot disassembly module for the part according to the part loss situation.

[0152] 2. In the intelligent disassembly platform for waste motor vehicles, compare the weight of the parts of the same category with the weight during the detection by the artificial intelligence acquisition module through the loss unit to judge the change in weight during the disassembly of the parts, and estimate the weight of the rust on the surface of the parts. Then, according to the data analysis unit records, analyzes the data of the part shape change, weight change and the weight change caused by the rust on the part surface. At this time, the recycling value comparison unit determines the recycling value threshold of the part according to the above data and judges the recycling value of the part. Brief Description of the Drawings

[0153] Figure 1 It is the overall module schematic diagram of the present invention;

[0154] Figure 2 It is the working principle flow chart of the present invention.

[0155] The meanings of the various labels in the figure are as follows:

[0156] 1. Motor vehicle transportation module; 100. Artificial intelligence acquisition module; 110. Image acquisition unit; 120. Weight acquisition unit; 130. Data update unit; 140. Part judgment unit; 200. Robot disassembly module; 210. Part grasping unit; 220. Stable signal input unit; 230. Operating trajectory modification unit; 240. Air purification unit; 300. Automated storage module; 310. Classification unit; 320. Part storage unit; 330. Loss unit; 340. Logistics transportation unit; 400. User interaction interface module; 410. Data analysis unit; 420. Recycling value comparison unit; 430. Display unit; Detailed implementation manners

[0157] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0158] An intelligent disassembly platform for waste motor vehicles includes a motor vehicle transportation module 1, an artificial intelligence acquisition module 100, a robot disassembly module 200, an automated storage module 300, and a user interaction interface module 400;

[0159] The motor vehicle transportation module 1 is used to transport the motor vehicle by an automatic guided vehicle;

[0160] The automatic guided vehicle automatically travels according to a preset route, automatically guides and scans the surrounding environment through a lidar to construct a three-dimensional map, detects obstacles, identifies ground markings or specific objects through a camera, and performs short-distance ranging with an ultrasonic sensor to detect obstacles in front. It matches the sensor data with a pre-established map to determine its own position, plans an optimal path based on the current position and the destination, and controls the steering and speed of the AGV according to the instructions provided by the navigation system.

[0161] The artificial intelligence acquisition module 100 includes an image acquisition unit 110, a weight acquisition unit 120, a data update unit 130, and a part judgment unit 140;

[0162] The image acquisition unit 110 is used to identify motor vehicle parts and the positions where the parts are located through multiple cameras, and output part position signals to the robot disassembly module 200;

[0163] The method for identifying motor vehicle parts is as follows:

[0164] After the camera identifies the motor vehicle parts, it compares the information of the motor vehicle parts identified by the camera with the information of the parts in the parts database. After the similarity is greater than 80%, the parts are determined, where the database is referenced by the parts data of the motor vehicle manufacturer;

[0165] First, from the feature vector of the part image to be identified and the reference part image in the parts database , the feature vector contains descriptive information such as color, texture, and shape;

[0166] Calculate the distance D between the two feature vectors. The calculation formula is as follows:

[0167] ;

[0168] Convert the feature distance to similarity. The similarity calculation formula is:

[0169]

[0170] is the maximum possible distance defined in advance, which is set to the sum of the maximum values of each dimension of the feature vector;

[0171] Compare the calculated similarity S with the preset threshold T. The preset threshold T is 80%. The comparison method is as follows:

[0172] If it is determined to be a similar part if , it is determined to be a similar part;

[0173] The method for identifying the position of the part is as follows:

[0174] Suppose there are two cameras, whose internal and external parameters are known. The internal parameters include the focal length and the principal point coordinates , and the external parameters include the rotation matrix (R) and the translation vector (T);

[0175] For two cameras, if they capture images of the same part, then the projection points of the part on the image planes of the two cameras are respectively and , and the three-dimensional coordinates of the part can be calculated through the following steps;

[0176] Calculate the projection matrix: The projection matrix P of each camera can be calculated from the internal and external parameters:

[0177]

[0178] where K is the internal parameter matrix and [R][T] is the external parameter matrix;

[0179] Establish a system of equations: Based on the projection matrix, establish the following system of equations to solve for the three-dimensional coordinates:

[0180] ;

[0181] By solving this system of equations, the three-dimensional coordinates of the part can be obtained ;

[0182] Convert the camera coordinate system to the actual world coordinate system. The calculation formula is as follows:

[0183] ;

[0184] Among them, is the actual world coordinate, is the coordinate in the camera coordinate system, is the rotation matrix, is the translation vector;

[0185] The weight acquisition unit 120 is used to measure the size of the motor vehicle part through laser ranging technology, and estimate the weight of the part by combining data such as material density;

[0186] After the image acquisition unit 110 acquires the information of the motor vehicle part, the weight acquisition unit 120 weighs the part. If the part weight does not correspond to the part threshold, it means that the internal components of the part are missing, and the part loss signal is output to the robot disassembly module 200;

[0187] Use a laser rangefinder to accurately measure the key dimensions of the motor vehicle part. The laser rangefinder determines the distance by emitting a laser pulse and measuring the time it takes to reflect back. Calculate the geometric shape of the part based on the measured dimension data. According to the geometric shape and dimension data of the part, use the corresponding geometric formula to calculate the volume of the part. Query the density of the material used for the part from the database, and calculate the weight of the part through the formula;

[0188] Volume calculation formula:

[0189] For regular shapes, the calculation formula for the volume V is as follows:

[0190]

[0191]

[0192]

[0193] Weight estimation formula:

[0194] ;

[0195] W is the weight of the part, V is the volume of the part, and p is the density of the material;

[0196] Example

[0197] Suppose we measure the dimensions of a motor vehicle part using laser ranging technology as follows:

[0198] ;

[0199] ;

[0200] ;

[0201] ;

[0202] Calculate the volume:

[0203] ;

[0204] Calculate the weight:

[0205] ;

[0206] Therefore, the estimated weight of the motor vehicle part is 234 kg;

[0207] The method for comparing the part weight with the part threshold is as follows:

[0208] Calculate the weight difference:

[0209] ;

[0210] Among them, is the weight difference, is the actually measured part weight, is the standard weight threshold preset through the manufacturer's part data;

[0211] If is greater than or equal to a preset error threshold , it is considered that the internal components of the part are missing;

[0212] ;

[0213] Suppose the weight of a standard part is 1.5 kg and the weight difference threshold set by the system is 0.1 kg; if during the weighing process, the actually measured weight is 1.3 kg, then the weight difference is 0.2 kg, which exceeds the threshold , and the system will determine that the internal components of the part are missing;

[0214] The data update unit 130 is used to collect and compare the weights of motor vehicle parts through the weight collection unit 120 when the image collection unit 110 fails to collect part information. After the comparison is completed, the part weight information is judged. According to the part weight information, the part shape information is distinguished respectively. After the distinction is completed, the part shape signal at this time is updated to the database in the image collection unit 110;

[0215] Based on the weight information, the following formula may be used to infer the part shape information: , where f is a function for inferring shape based on weight, and the technical formula of this function is as follows:

[0216]

[0217] Where, , ,…, Regression coefficient, is the error term

[0218] Working principle:

[0219] When a new part weight is measured, we use the above formula to predict its shape characteristics ;

[0220] By solving the equation = to infer the shape characteristics;

[0221] The shape characteristics can then be used to determine the shape of the part;

[0222] If the inference is successful, the part information in the database is updated: number;

[0223] The part judgment unit 140 is used to judge whether a motor vehicle part is missing according to the image collection unit 110, and output the part missing signal to the robot disassembly module 200;

[0224] If a part is missing, it means that both the image collection unit 110 and the weight collection unit 120 have detected part information;

[0225] Working principle: The image collection unit 110 obtains image information by photographing relevant parts of the motor vehicle, analyzes and processes the image to judge whether a part exists, and the weight collection unit 120 measures the weight of relevant parts of the motor vehicle. The part judgment unit 140 comprehensively uses the information provided by these two units to determine whether a part is missing;

[0226] If a part is missing, it means that the image acquisition unit 110 fails to detect the features of a specific part in the image it acquires, and at the same time, there is a difference between the weight measured by the weight acquisition unit 120 and the weight that should be at this part in the complete state;

[0227] Calculation formula: Assume that the standard weight of a certain part of a complete motor vehicle is W standard, the number of parts that the image acquisition unit 110 can identify is N image, the actual weight measured by the weight acquisition unit 120 is W actual, and the average weight of each part is W average;

[0228] Then it is possible to judge whether a part is missing through the following formula: If N image is less than the expected number of parts, and , then it is determined that the part is missing.

[0229] The robot disassembly module 200 includes a part grasping unit 210, a stable signal input unit 220, and a running trajectory modification unit 230;

[0230] The part grasping unit 210 is used to classify and disassemble the motor vehicle parts through an intelligent gripper according to the shape, color, and size of the parts identified by the artificial intelligence acquisition module 100;

[0231] Distance calculation: ;

[0232] Angle calculation: ;

[0233] Among them and are the starting point and ending point coordinates respectively;

[0234] Grasping force calculation:

[0235] ;

[0236] Among them, F is the grasping force, m is the part mass, and g is the acceleration due to gravity;

[0237] The stable signal input unit 220 is used to stably receive the part position signal, part loss signal, and part missing signal output from the artificial intelligence acquisition module 100 through radio frequency;

[0238] The radio frequency alarm signal generates a radio wave signal containing alarm information through the transmitter in the alarm system. This signal is generated by an oscillator and transmitted through an antenna; the radio wave propagates in the air and penetrates walls and other obstacles; the receiving antenna of the power output module captures the radio wave and transmits it to the internal receiving circuit, and the receiving circuit decodes the received radio wave signal to identify the alarm information;

[0239] The running trajectory modification unit 230 is used to modify the running trajectory of the part grasping unit 210 according to the signal of the stability signal input unit 220, where the running trajectory includes the grasping position and clamping force of the mechanical gripper;

[0240] An example of the calculation formula is as follows:

[0241] Assume the initial grasping position is , according to the signal of the stability signal input unit 220, the fine-tuning calculation formula is as follows:

[0242] , ,

[0243] The adjusted signal is: ;

[0244] where, , , is the signal input by the stability signal input unit 220;

[0245] Adjustment of the clamping force:

[0246] Assume the part weight changes from to , and the corresponding clamping force changes from to , then the clamping force change value Formula calculation:

[0247] ;

[0248] where:

[0249] K is the proportionality coefficient of the clamping force to the weight,

[0250] is the weight of the part after change;

[0251] is the reference weight;

[0252] Example:

[0253] Assume the reference weight , and the corresponding reference clamping force , the proportionality coefficient

[0254] If the part weight increases to , then the clamping force change value is:

[0255] ;

[0256] Therefore, the clamping force needs to be increased by 5 N to adapt to the increase in the weight of the parts.

[0257] The robot disassembly module 200 also includes an air purification unit 240 for treating harmful substances generated during the classification and disassembly process of the part grasping unit 210 by the air filtration method;

[0258] The working principle of the air filtration method is to filter and adsorb harmful substances in the air through a filter, so as to achieve the purpose of purifying the air:

[0259] The effect of the air filtration method is estimated by the following formula:

[0260] ;

[0261] Among them, the filtration efficiency represents the ability of the filter to remove harmful substances.

[0262] The automated storage module 300 includes a classification unit 310, a part storage unit 320, and a loss unit 330;

[0263] The classification unit 310 is used to classify the parts disassembled by the robot disassembly module 200, and the classification threshold is calculated based on the combination of the shape and weight of the parts in the artificial intelligence acquisition module 100;

[0264] Analyze the average values of the part shapes and weights in the database of the artificial intelligence acquisition module 100;

[0265] Comprehensively evaluate the part classification threshold according to the average estimates of weight and shape. Since the part shape and weight data in the database are updated in real time, the part classification threshold changes in real time according to the changes in the database;

[0266] Classification threshold calculation formula:

[0267] Suppose we have n parts, and each part has a shape feature and a weight feature ;

[0268] Calculate the average values of shape and weight:

[0269] ;

[0270] ;

[0271] Suppose we use the linear weighting method to calculate the comprehensive score , where and are the weights of shape and weight, both of which are 0.5:

[0272] ;

[0273] is the classification threshold;

[0274] The part storage unit 320 is used to store the parts classified by the classification unit 310, and the same parts are divided into different qualities;

[0275] According to the comprehensive score of the classification unit 310 Set the classification threshold, and the quality discrimination method is as follows:

[0276] is excellent;

[0277] is poor;

[0278] Among them, and are the comprehensive scores of the parts;

[0279] The loss unit 330 is used to compare the weight of the parts stored in the same category in the part storage unit 320 with the weight during the detection by the artificial intelligence acquisition module 100, judge the change in weight during the disassembly process of the parts, infer the weight of the rust on the surface of the parts, and then output the rust change signal to the user interface module 400;

[0280] The weight calculation formula is as follows:

[0281] ;

[0282] is the total weight of the parts in the same category in the storage unit, is the weight during the detection by the artificial intelligence acquisition module 100. By analyzing the change trend of, infer the rust situation.

[0283] The automated storage module 300 further includes a logistics transportation unit 340. The logistics transportation unit 340 is used to automatically notify the logistics platform for cleaning by radio frequency when the part storage unit 320 is full.

[0284] The user interface module 400 includes a data analysis unit 410 and a recycling value comparison unit 420;

[0285] The data analysis unit 410 is used to record and analyze the data on the shape change, weight change of the parts and the weight change caused by the rust on the surface of the parts, and determine the recycling value threshold of the motor vehicle parts according to the analyzed data;

[0286] Assume that the original value of the part is , then the recycling value threshold V is calculated by the following formula:

[0287] ;

[0288] a is the shape change coefficient, is the weight change coefficient, and r is the rust weight change coefficient;

[0289] The recycling value comparison unit 420 is used to compare the recycling value and determine the recycling value threshold of motor vehicle parts according to the data analysis unit 410;

[0290] The recycling value is 60% of the recycling value threshold, and the classification unit 310 calculates 40% of the classification threshold;

[0291] If , then the recycling value of this part is high;

[0292] If , then the recycling value of this part is low.

[0293] The user interaction interface module 400 further includes a display unit 430 for sorting the recycling values of motor vehicle parts in the recycling value comparison unit 420 from large to small and remotely outputting them to the staff via radio frequency.

[0294] The image acquisition unit 110 acquires the image features of waste motor vehicle parts, determines the position and part signal of the motor vehicle parts according to the image features. When the recognition of the image acquisition unit 110 is inaccurate, the weight acquisition unit 120 determines the part type according to the weight of the motor vehicle parts, acquires the part shape, and updates the data of the image acquisition unit 110;

[0295] When the part threshold collected by the weight acquisition unit 120 does not correspond to the part weight, if the image acquisition unit 110 acquires the information of the motor vehicle part, it means that the internal components of the part are missing;

[0296] When the image acquisition unit 110 cannot distinguish the part by shape and weight, it means that the part is missing;

[0297] The stable signal input unit 220 modifies the operation trajectory of the part grasping unit 210 through the part position signal, part loss signal, and part missing signal;

[0298] The weight of the part in the part storage unit 320 is compared with the weight of the artificial intelligence acquisition module 100 to judge the change in the part weight. The recycling value comparison unit 420 judges the recycling value of the part, and the threshold of the recycling value is judged by modifying the data of the robot disassembly module 200 and the weight change data in the loss unit 330.

[0299] The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and the descriptions in the specification are only preferred examples of the present invention, which are not used to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of the present invention claimed is defined by the appended claims and their equivalents.

Claims

1. An intelligent dismantling platform for scrapped motor vehicles, characterized by: It comprises a motor vehicle transport module (1), an artificial intelligence collection module (100), a robot disassembly module (200), an automated storage module (300) and a user interaction interface module (400); The artificial intelligence acquisition module (100) comprises an image acquisition unit (110), a weight acquisition unit (120), a data update unit (130) and a part judgment unit (140); the robot disassembly module (200) comprises a part grasping unit (210), a stable signal input unit (220) and an operation track modification unit (230); the automated storage module (300) comprises a classification unit (310), a part storage unit (320) and a loss unit (330); and the user interaction interface module (400) comprises a data analysis unit (410) and a recycling value comparison unit (420); The image acquisition unit (110) acquires image features of waste motor vehicle parts, and determines the position of motor vehicle parts and part signals based on the image features. When the image acquisition unit (110) makes inaccurate identification, the weight acquisition unit (120) determines the type of parts based on the weight of the motor vehicle parts, and acquires the shape of the parts to update the data of the image acquisition unit (110). When the weight acquisition unit (120) acquires part threshold values ​​that do not correspond to part weights, if the image acquisition unit (110) acquires motor vehicle part information, it indicates that the internal components of the parts are missing. When the image acquisition unit (110) cannot distinguish the parts by shape and weight, it indicates that the parts are missing. The stable signal input unit (220) modifies the running track of the part grabbing unit (210) through the part position signal, the part loss signal and the part missing signal; The weight of the parts in the parts storage unit (320) is compared with the weight of the artificial intelligence acquisition module (100) to determine the change in the weight of the parts, and the recycling value comparison unit (420) determines the recycling value of the parts, and the recycling value threshold is determined by modifying the data of the robot disassembly module (200) and the weight change data in the loss unit (330); The part grabbing unit (210) is used to classify and split motor vehicle parts through an intelligent gripper according to the shape, color and size of the parts identified by the artificial intelligence acquisition module (100); The stable signal input unit (220) is used to stably receive the part position signal, the part loss signal and the part missing signal outputted from the artificial intelligence acquisition module (100) via radio frequency; The running trajectory modification unit (230) is used to modify the running trajectory of the part grasping unit (210) according to the signal of the stable signal input unit (220), wherein the running trajectory includes the grasping position and clamping force of the mechanical gripper.

2. The intelligent dismantling platform for scrapped motor vehicles according to claim 1 is characterized by: The motor vehicle transport module (1) is used to transport a motor vehicle by means of an automatic guided vehicle.

3. The intelligent dismantling platform for scrapped motor vehicles according to claim 1 is characterized in that: The image acquisition unit (110) is used to identify motor vehicle parts and the positions of the parts through a plurality of cameras, and output part position signals to the robot disassembly module (200); The weight collection unit (120) is used to measure the size of the motor vehicle part by laser ranging technology and estimate the weight of the part in combination with material density data; After the image acquisition unit (110) acquires the motor vehicle part information, the weight acquisition unit (120) weighs the part; if the part weight does not correspond to the part threshold, it indicates that the internal component of the part is missing, and a part missing signal is output to the robot disassembly module (200); The data updating unit (130) is used to collect and compare the weight of the motor vehicle parts through the weight collection unit (120) when the image collection unit (110) cannot collect the part information, determine the part weight information after the comparison is completed, distinguish the part shape information according to the part weight information, and update the part shape signal at this time to the database in the image collection unit (110) after the distinction is completed; The part judgment unit (140) is used to judge whether a motor vehicle part is missing based on the image acquisition unit (110), and output a part missing signal to the robot disassembly module (200).

4. The intelligent dismantling platform for scrapped motor vehicles according to claim 1 is characterized by: The classification unit (310) is used to classify the parts disassembled by the robot disassembly module (200), and the classification threshold is calculated based on the shape and weight of the parts in the artificial intelligence acquisition module (100); Analyzing the average value of the shape and weight of the parts in the database of the artificial intelligence acquisition module (100); The part classification threshold is comprehensively judged based on the average weight and shape estimates. Since the part shape and weight data in the database are updated in real time, the part classification threshold changes in real time according to the changes in the database. The parts storage unit (320) is used to store the parts classified by the classification unit (310), and the same parts are classified into different qualities, and the classification threshold is calculated based on the shape and weight of the parts in the artificial intelligence acquisition module (100); The loss unit (330) is used to compare the weight of the same category of parts stored in the parts storage unit (320) with the weight detected by the artificial intelligence acquisition module (100), determine the change in weight during the disassembly of the parts, infer the weight of rust on the surface of the parts, and then output a rust change signal to the user interaction interface module (400).

5. The intelligent dismantling platform for scrapped motor vehicles according to claim 1 is characterized by: The data analysis unit (410) is used to record and analyze data on shape changes, weight changes, and weight changes caused by rust on the surface of parts, and determine a threshold value for recycling the motor vehicle parts based on the analyzed data; The recycling value comparison unit (420) is used for recycling value comparison to determine the recycling value threshold of motor vehicle parts according to the data analysis unit (410).

6. The intelligent dismantling platform for scrapped motor vehicles according to claim 1 is characterized by: The robot disassembly module (200) further comprises an air purification unit (240) for treating harmful substances generated during the classification and disassembly process of the part grabbing unit (210) through an air filtration method.

7. The intelligent dismantling platform for scrapped motor vehicles according to claim 1 is characterized by: The automated storage module (300) further comprises a logistics transport unit (340), wherein the logistics transport unit (340) is used to automatically notify the logistics platform to clear the parts storage unit (320) via radio frequency when the parts storage unit (320) is fully loaded.

8. The intelligent dismantling platform for scrapped motor vehicles according to claim 1 is characterized by: The user interaction interface module (400) further comprises a display unit (430) for sorting the recycling values ​​of the motor vehicle parts in the recycling value comparison unit (420) from large to small, and remotely outputting the sorting to a staff member via a radio frequency.

Images