High-safety mechanical garage intelligent parking method and system

Through the garage center server and sensor system, vehicle information is intelligently evaluated, suitable parking spaces are determined and vehicles are parked through mechanical equipment, which solves the problem of poor management of mechanical three-dimensional garages and achieves efficient and safe parking services and parking space utilization.

CN119091673AInactive Publication Date: 2024-12-06CHONGQING SPECIAL EQUIP INSPECTION & RES INST
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411182052.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-06
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Poor management of mechanical three-dimensional garages leads to inability to guide vehicles in time, long parking process, low garage utilization rate, poor user experience, how to improve the parking safety and utilization rate of mechanical garages.

Method used

Receive user's car storage and car collection information through the garage center server, use weighing sensors and infrared sensors to identify the vehicle, build an optimal driving path model, intelligently evaluate the vehicle's weight and size information, determine the appropriate parking space, and use mechanical equipment to Park the vehicle safely to ensure evenly distributed loads and avoid overload.

Benefits of technology

It realizes intelligent parking services with high safety, improves parking efficiency and user experience, ensures safe parking of vehicles, optimizes the utilization rate of parking spaces, reduces waiting time, and prevents overload.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119091673A_ABST
    Figure CN119091673A_ABST
Patent Text Reader

Abstract

The invention discloses a high-safety mechanical garage intelligent parking method and system, and relates to the technical field of intelligent parking. When a vehicle arrives at a designated place, the vehicle is identified; guiding the user to enter the parking transition area; a weighing sensor is buried below the parking transition area, after the vehicle enters the parking transition area, the weight information of the vehicle is obtained through the weighing sensor, and the bearing capacity of each parking space is judged according to the weight information of the vehicle; an infrared sensor is installed at the entrance or the periphery of the parking transition area to obtain the size information of the vehicle; the weight information and the size information of the vehicle are intelligently evaluated, an available parking space suitable for the vehicle is determined, and the vehicle is parked in the available parking space through mechanical equipment; and the vehicle is moved from the parking space to the parking transition area through mechanical equipment. According to the invention, accurate monitoring of the real-time weight of each parking space is realized, and the use state of the parking space is dynamically analyzed, so that uniform load distribution is ensured, and the safety of the mechanical parking garage is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent parking, and more particularly to a highly secure intelligent parking method and system for a mechanical garage. Background Art

[0002] With the acceleration of urbanization, the number of cars has continued to increase, and parking difficulties and high parking costs have become a problem for many car owners. In order to solve this problem, mechanical stereo garages have emerged. They use stereo space to achieve efficient use of parking spaces, while also saving land resources and reducing environmental pollution.

[0003] However, although the prospects for multi-story parking lots are good, there are also problems such as poor management that result in vehicles not being evacuated in a timely manner, a long parking process, low garage utilization, and poor user experience.

[0004] Therefore, how to propose a highly secure intelligent parking method and system for a mechanical garage to realize intelligent parking and car retrieval services while improving the parking safety of the mechanical garage is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0005] In view of this, the present invention provides a highly secure intelligent parking method and system for a mechanical garage, which realizes intelligent parking and car retrieval services and improves the parking safety of the mechanical garage.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a highly safe mechanical garage intelligent parking method, comprising:

[0007] The garage central server receives parking information or pick-up information provided by the user who wants to park; the parking information includes the user's license plate number, parking arrival time and estimated parking duration information; the user reserves a parking space through the mobile terminal and sends the license plate number, parking arrival time and estimated parking duration information to the garage central server at the same time; the user can make a reservation in advance at home or on the way to the garage; the pick-up information includes the user's pick-up arrival time; helping the garage central server to effectively arrange parking spaces and manage parking processes;

[0008] When the vehicle arrives at the designated location, it is identified based on the license plate information to ensure that only authorized vehicles can enter;

[0009] Build an optimal driving path model, and after identification, plan the optimal driving path for the user and guide the user into the parking transition area;

[0010] A weighing sensor is buried under the parking transition area. After a vehicle enters the parking transition area, the weighing sensor obtains the weight information of the vehicle, and the load-bearing capacity of each parking space is determined according to the weight information of the vehicle;

[0011] Install infrared sensors at the entrance or around the parking transition area to obtain the size information of the vehicle;

[0012] Intelligently evaluate the weight and size information of the vehicle, determine the available parking space suitable for the vehicle, and park the vehicle safely in the available parking space through mechanical equipment;

[0013] According to the vehicle pick-up information, the vehicle is moved from the parking space to the parking transition area by mechanical equipment.

[0014] Preferably, intelligently evaluating the weight information and size information of the vehicle to determine an available parking space suitable for the vehicle; and safely parking the vehicle in the available parking space by mechanical equipment, including:

[0015] Set a maximum load-bearing threshold for each parking space, analyze the weight data of each parking space based on an intelligent algorithm, and automatically mark a vehicle as unavailable when its weight reaches or exceeds the threshold;

[0016] The intelligent algorithm analyzes the maximum load-bearing capacity of each parking space based on the current vehicle weight data, and evaluates the stress condition of the entire mechanical parking garage in combination with the weight data of all parking spaces;

[0017] Set the maximum size threshold for each parking space, compare and analyze the size information of each parking space, and automatically mark a vehicle as unavailable when its size reaches or exceeds the threshold.

[0018] When the vehicle's weight information and size information both meet the above conditions, it is determined to be an available parking space;

[0019] Move the vehicle to an available parking space.

[0020] Preferably, the method further comprises: installing a weighing sensor on each parking space to obtain real-time weight data of each parking space;

[0021] Update the usage status of each parking space in real time.

[0022] Preferably, the intelligent algorithm analyzes the maximum load-bearing capacity of each parking space based on the weight data of the current vehicle, and evaluates the stress condition of the entire mechanical parking garage in combination with the weight data of all parking spaces, including:

[0023] Establish a three-dimensional coordinate system for the garage, where (x, y, z) represents the actual coordinate center position of the parking space;

[0024] Create parking space matrix R i(x, y, z), i represents the number of the parking space, x represents the x-axis coordinate of the i-th parking space, y represents the y-axis coordinate of the i-th parking space, and z represents the z-axis coordinate of the i-th parking space;

[0025] Initialize n smart matching operators. The smart matching operators work in a cyclic calculation mode. Set the total number of cyclic calculations to c. Each smart matching operator can perform matching calculations and position change calculation tasks. Establish the smart matching operator matrix M. k (x'k(j), y'k(j), z'k(j), P), k represents the number of the smart matching operator, x'k(j) represents the x-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, y'k(j) represents the y-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, z'k(j) represents the z-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, and P represents the vehicle information loaded by the k-th smart matching operator; the information of available parking spaces in the garage is loaded into the smart matching operator matrix, and the random number generation algorithm is used to randomly assign x, y, and z coordinates to each smart matching operator, thereby initializing the smart matching operator to the three-dimensional coordinate system of the garage;

[0026] The initialized smart matching operator performs a matching calculation if there is an available parking space at the current initialization position. If the matching result meets the maximum load-bearing and maximum accommodating size thresholds of the parking space, it is considered a successful match and the matching result Sui is recorded in the matching history information table His;

[0027] If the matching result does not meet the maximum load-bearing and / or maximum accommodating size thresholds of the parking space, the matching is unsuccessful, and the position information will be calculated using the smart matching operator displacement algorithm based on the conversion switch in the next cycle of the smart matching operator.

[0028] Preferably, the method further includes: using encryption technology during data transmission and storage to encrypt and protect the information and parking records provided by the user, thereby ensuring the security of the user information and parking records and protecting the user's privacy.

[0029] Preferably, it also includes: setting up a safety alarm device to promptly alarm and activate the emergency rescue mechanism when an emergency occurs.

[0030] Preferably, an optimal driving path model is constructed, and after identification, an optimal driving path is planned for the user, including:

[0031] Get the endpoint position of the vehicle and the endpoint position of the parking transition area,

[0032] Calculate the best ordered inflection point combination between the vehicle position and the parking transition area according to the relationship between the endpoint position of the vehicle and the endpoint position of the parking transition area;

[0033] The first endpoint information in the best ordered inflection point combination is extracted and sent to the user's mobile terminal; when the endpoint position of the vehicle changes, the optimal driving path is recalculated.

[0034] Preferably, the vehicle is safely parked in a suitable parking space by mechanical equipment based on the improved Mask-RCNN network model, including:

[0035] Collect images of the parking status of various vehicles in parking spaces, and divide them into sample data sets and data of vehicles to be parked;

[0036] Input the sample data set and the data of the vehicles to be parked into the CNN network of the improved Mask-RCNN network model to extract image features;

[0037] The feature map is obtained by integrating image features through the improved EC module; the ROI of the feature map is extracted and filtered through the RPN network;

[0038] The region of interest alignment layer is used to unify the size of the feature map according to the filtered ROI. Finally, the feature map is subjected to binary classification, bounding box regression, and mask segmentation in the fully connected layer to generate a binary mask image, and a multi-task loss function L of the binary mask image is established.

[0039] The value of the multi-task loss function L is reduced by continuously adjusting the parameters of binary classification, bounding box regression, and Mask until the global optimal solution of the multi-task loss function L is obtained;

[0040] The global optimal solution and weight matrix of the multi-task loss function L are substituted into the vehicle safe parking function, and the Mask-RCNN network model parameters are improved through multiple iterations until the optimal solution of the vehicle safe parking function is obtained; according to the optimal solution, the mechanical equipment is controlled to safely park the vehicle in an available parking space.

[0041] Preferably, the weight matrix is ​​expressed as:

[0042] ω l =[d1, d2…d i …d n ];

[0043] n is the number of sample data sets and vehicle data to be parked,

[0044] d i The expression is as follows:

[0045]

[0046] In the above formula, α C , α S , αH are the actual contributions of the color feature vector, texture feature vector, and shape feature vector of the parking state of the vehicle in the parking space to the similarity measurement; are the actual contributions of the total similarity measure of the N1 color feature vector, N2 texture feature vector, and N3 shape feature of the parking state of the vehicle in the parking space, are the color feature vector, texture feature vector, and shape feature vector of the i-th image of the vehicle to be parked in the parking space, respectively. are respectively the color feature vector, texture feature vector, and shape feature vector of the i-th image in the sample data set of the vehicle to be parked in the parking state in the parking space, P is the vehicle data to be parked, T is the sample data set, and i is the time sequence number of the image in the data set.

[0047] Preferably, the expression of the vehicle safe parking function f(x) is as follows:

[0048]

[0049] In the formula, L is the loss function, H l is the activation function, x i is the vehicle to be parked, ω l is the weight matrix, y i is the grayscale image function, and n is the number of sample data sets and vehicle data to be parked.

[0050] The optimal solution of the vehicle safety parking function f(x) is expressed as follows:

[0051]

[0052] In the above formula, x i is a sample data set, H l ' is the activation function H l The partial derivative of the weight matrix, η is the regularization coefficient.

[0053] Preferably, a path planning algorithm is used to calculate the optimal driving path, including:

[0054] Determine the endpoint position of the vehicle, the endpoint position of the parking transition area, and the time when the user arrives at the garage;

[0055] Obtain all available driving paths based on the endpoint position of the vehicle and the endpoint position of the parking transition area; calculate the distance, time and congested vehicles ahead of each driving path;

[0056] An initial planning path is obtained from all driving paths according to the saving algorithm; the inflection point closest to the vehicle's endpoint is selected as the first inflection point, the estimated time from other inflection points to the next inflection point and the estimated time from the first inflection point to the next inflection point are calculated, and the time difference is evaluated according to the saving algorithm to determine whether to use other inflection points as selected inflection points until the iteration condition is met, and all selected inflection points are arranged in chronological order to form an initial planning path;

[0057] A large number of initial populations are generated from all the initial planning paths obtained, and the optimal driving path is obtained by iteration combined with genetic algorithm.

[0058] Preferably, a highly secure mechanical stereo garage intelligent parking system comprises:

[0059] The information receiving module is used to receive parking information or pick-up information provided by the parking user through the garage central server; the parking information includes the user's license plate number, parking arrival time and estimated parking duration information; the pick-up information includes the user's pick-up arrival time;

[0060] The vehicle identification module is used to identify the vehicle according to the license plate number information when the vehicle arrives at the designated location;

[0061] The path planning module is used to build the optimal driving path model. After identification, it plans the optimal driving path for the user and guides the user into the parking transition area;

[0062] A weight information acquisition module is used to bury a weighing sensor under the parking transition area. After a vehicle enters the parking transition area, the weighing sensor acquires the weight information of the vehicle and determines the load-bearing capacity of each parking space according to the weight information of the vehicle;

[0063] A size information acquisition module is used to install an infrared sensor at the entrance or around the parking transition area to acquire the size information of the vehicle;

[0064] A vehicle parking module, which is used to intelligently evaluate the weight and size information of the vehicle, determine an available parking space suitable for the vehicle, and safely park the vehicle in the available parking space through mechanical equipment;

[0065] The vehicle extraction module is used to move the vehicle from the parking space to the parking transition area through mechanical equipment according to the vehicle extraction information.

[0066] Through the above technical solutions, it can be known that compared with the prior art, the present invention discloses a highly secure mechanical stereo garage intelligent parking method and system, including: receiving parking information or pick-up information provided by a user to be parked through a garage central server; the parking information includes the user's license plate number, parking arrival time and expected parking time information; the pick-up information includes the user's pick-up arrival time; when the vehicle arrives at the designated location, the vehicle is identified according to the license plate number information; an optimal driving path model is constructed, and after the identification is passed, the optimal driving path is planned for the user to guide the user to enter the parking transition area; a weighing sensor is buried under the parking transition area, and after the vehicle enters the parking transition area, the weight information of the vehicle is obtained through the weighing sensor, and the load-bearing capacity of each parking space is determined according to the weight information of the vehicle; an infrared sensor is installed at the entrance or around the parking transition area to obtain the size information of the vehicle; the weight information and size information of the vehicle are intelligently evaluated to determine the available parking space suitable for the vehicle, and the vehicle is safely parked in the available parking space through mechanical equipment; according to the pick-up information, the vehicle is moved from the parking space to the parking transition area through mechanical equipment. In the present invention, the parking and retrieval information provided by the user is processed by the central server, and the system provides the user with the optimal driving path and guides the user to park the vehicle. The entire process is carried out by mechanical equipment, which ensures safety, improves parking efficiency, and ensures that the vehicle is parked safely. It can provide efficient parking services and optimize user experience. By reserving parking spaces in advance and intelligent navigation, users can park and retrieve vehicles more conveniently, reducing waiting time during parking. Based on the intelligent algorithm, intelligent evaluation is performed according to the weight information of each vehicle to achieve the optimal state of system stress and effectively predict the use efficiency of parking spaces. Ensure that the load is evenly distributed to avoid overload. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0068] Figure 1 A schematic flow chart of a highly secure intelligent parking method for a mechanical garage provided by the present invention.

[0069] Figure 2 A schematic structural diagram of a highly secure mechanical garage intelligent parking system provided by the present invention. DETAILED DESCRIPTION

[0070] 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 only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] The embodiment of the present invention discloses a highly secure mechanical stereo garage intelligent parking method and system. The parking information and vehicle retrieval information provided by the user are processed by a central server. The system provides the user with an optimal driving path and guides the user to park the vehicle. The entire process is carried out by mechanical equipment, which ensures safety, improves parking efficiency, and ensures safe parking of vehicles.

[0072] In a specific embodiment of the present invention, a highly secure mechanical garage intelligent parking method, such as Figure 1 As shown, including:

[0073] The garage central server receives parking information or pick-up information provided by the user who wants to park; the parking information includes the user's license plate number, parking arrival time and estimated parking duration information; the user reserves a parking space through the mobile terminal and sends the license plate number, parking arrival time and estimated parking duration information to the garage central server at the same time; the user can make a reservation in advance at home or on the way to the garage; the pick-up information includes the user's pick-up arrival time; helping the garage central server to effectively arrange parking spaces and manage parking processes;

[0074] When the vehicle arrives at the designated location, it is identified based on the license plate information to ensure that only authorized vehicles can enter;

[0075] Build an optimal driving path model, and after identification, plan the optimal driving path for the user and guide the user into the parking transition area;

[0076] A weighing sensor is buried under the parking transition area. After a vehicle enters the parking transition area, the weighing sensor obtains the weight information of the vehicle, and the load-bearing capacity of each parking space is determined according to the weight information of the vehicle, thereby performing an assessment;

[0077] Install infrared sensors at the entrance or around the parking transition area to detect the vehicle's entry and exit status and outline size. Obtain vehicle size information to assist in parking space allocation;

[0078] Intelligently evaluate the weight and size information of the vehicle, determine the available parking space suitable for the vehicle, and park the vehicle safely in the available parking space through mechanical equipment;

[0079] According to the vehicle pick-up information, the vehicle is moved from the parking space to the parking transition area by mechanical equipment.

[0080] Users can directly go to the parking transition area to pick up their cars, reducing waiting time. Similarly, after users pick up their cars, the optimal driving path model can also be used to plan the optimal driving path for users, ensuring that users can exit the garage safely and quickly, effectively ensuring the safety of the parking process, improving user experience and reducing safety risks.

[0081] Specifically, the weight and size information of the vehicle are intelligently evaluated to determine an available parking space suitable for the vehicle; the vehicle is safely parked in the available parking space through mechanical equipment, including:

[0082] Set the maximum load-bearing threshold for each parking space, analyze the weight data of each parking space based on an intelligent algorithm, and automatically mark a vehicle as unavailable when its weight reaches or exceeds the threshold to prevent overloading;

[0083] The intelligent algorithm analyzes the maximum load-bearing capacity of each parking space based on the current vehicle weight data, and evaluates the stress condition of the entire mechanical parking garage in combination with the weight data of all parking spaces;

[0084] Set the maximum size threshold for each parking space, compare and analyze the size information of each parking space, and automatically mark a vehicle as unavailable when its size reaches or exceeds the threshold.

[0085] When the vehicle's weight information and size information both meet the above conditions, it is determined to be an available parking space;

[0086] Move the vehicle to an available parking space.

[0087] According to the parking space load information and size information, for heavier or large vehicles, the system can give priority to those parking spaces with lighter loads and stronger load-bearing capacity to ensure the safety and balance of the entire parking area. At the same time, according to the size of each vehicle, different parking spaces are adopted to make large vehicles park in large spaces and small vehicles park in small spaces.

[0088] Based on the intelligent algorithm, intelligent evaluation is performed according to the weight information of each vehicle to achieve the optimal stress state of the system and effectively predict the utilization efficiency of parking spaces. Ensure that the load is evenly distributed to avoid overload.

[0089] Specifically, it also includes: installing a weighing sensor on each parking space to obtain real-time weight data of each parking space; real-time monitoring of the weight change of each parking space; collecting data through sensors and updating the usage status of the system parking space in real time;

[0090] Update the usage status of each parking space in real time.

[0091] The use status of the parking space includes: available, occupied, and alarm.

[0092] Specifically, the intelligent algorithm analyzes the maximum load-bearing capacity of each parking space based on the current vehicle weight data, and evaluates the stress conditions of the entire mechanical parking garage in combination with the weight data of all parking spaces, including:

[0093] (1) Establish a three-dimensional coordinate system for the garage, where (x, y, z) represents the actual coordinate center position of the parking space;

[0094] (2) Establish parking space matrix R i (x, y, z), i represents the number of the parking space, x represents the x-axis coordinate of the i-th parking space, y represents the y-axis coordinate of the i-th parking space, and z represents the z-axis coordinate of the i-th parking space;

[0095] (3) Initialize n smart matching operators. The smart matching operators work in a cyclic calculation mode, and the total number of cyclic calculations is set to c. Each smart matching operator can perform matching calculation and position change calculation tasks, and establish the smart matching operator matrix M. k (x'k(j), y'k(j), z'k(j), P), k represents the number of the smart matching operator, x'k(j) represents the x-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, y'k(j) represents the y-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, z'k(j) represents the z-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, and P represents the vehicle information loaded by the k-th smart matching operator; the information of available parking spaces in the garage is loaded into the smart matching operator matrix, and the random number generation algorithm is used to randomly assign x, y, and z coordinates to each smart matching operator, thereby initializing the smart matching operator to the three-dimensional coordinate system of the garage;

[0096] (4) If there is an available parking space at the current initialization position, the initialized smart matching operator performs a matching calculation. If the matching result meets the maximum load-bearing and maximum accommodating size thresholds of the parking space, it is considered a successful match, and the matching result Sui is recorded in the matching history information table His;

[0097] If the matching result does not meet the maximum load-bearing and / or maximum accommodating size thresholds of the parking space, the matching is unsuccessful, and the position information will be calculated using the smart matching operator displacement algorithm based on the conversion switch in the next cycle of the smart matching operator.

[0098] Set the conversion switch Swi. The conversion switch Swi is used to determine the search strategy of the smart matching operator. When Swi is greater than 0.6, the search strategy of the smart matching operator is suitable for the local search strategy, that is, the smart matching operator generates new position information according to the smart matching operator displacement algorithm. When Swi is less than 0.6, the search strategy of the smart matching operator is suitable for the wide-area search strategy, that is, the smart matching operator generates new position information according to the random number generation algorithm. Assume that the conversion switch of the k-th smart matching operator in the j-th cycle is denoted by Swi j k , The calculation formula is as follows:

[0099]

[0100] in the formula is the maximum value of the matching degree Sui of all intelligent matching operators in the jth cycle; is the minimum value of the matching degree Sui of all intelligent matching operators in the jth cycle; j is the current cycle number, and c is the total number of cycles;

[0101] When Swi is greater than 0.6, the smart matching operator search strategy implements a local search strategy, and the displacement algorithm of the kth smart matching operator in the jth cycle is as follows:

[0102] X k '(j+1)=X k '(j)+|λ1 2 *V j -X k '(j)|*Ste;

[0103] Y k '(j+1)=Y k '(j)+|λ2 2 *V j -Y k '(j)|*Ste;

[0104] Z k '(j+1)=Z k '(j)+|λ3 2 *V j -Z k '(j)|*Ste;

[0105] X in the formula k '(j), Y k '(j), Z k '(j) are the x-axis coordinate, y-axis coordinate, and z-axis coordinate of the k-th smart matching operator in the j-th cycle; λ1, λ2, and λ3 are the first, second, and third random numbers, respectively, and their value range is [0, 1]; V jis the number of successful matching results in the whole system in the jth cycle; Ste is the distance between the kth smart matching operator and the nearest other smart matching operators in the jth cycle;

[0106] (5) After the intelligent matching operator generates a new position, if there is an available parking space at the current initialization position, a matching calculation is performed. If the matching result meets the maximum load-bearing and maximum accommodating size thresholds of the parking space, it is considered a successful match, and the matching result Sui is recorded in the matching history information table His.

[0107] By implementing an intelligent weighing monitoring system in a mechanical parking garage, it is possible to accurately monitor the real-time weight of each parking space and dynamically analyze the usage status of the parking space, thereby ensuring even distribution of the load and improving the safety and efficiency of the mechanical parking garage. The system can not only prevent overloading, but also effectively improve parking efficiency and reduce operational risks.

[0108] Specifically, it also includes: using encryption technology during data transmission and storage to encrypt and protect the information and parking records provided by users, ensuring the security of user information and parking records while protecting user privacy.

[0109] The encryption module of the user's mobile terminal encrypts the user information to obtain encrypted information; the central server decrypts the encrypted information and verifies the authenticity of the data sent by the user's mobile terminal; if the verification is successful, the corresponding control instructions are executed to complete the corresponding operation; if not, it is not executed;

[0110] The encryption module of the user mobile terminal encrypts the user information to obtain the encrypted information, specifically including: the user mobile terminal obtains the mixed encryption value by calculation according to the designated user number and the mixed encryption algorithm during encryption, and sends the mixed encryption value to the central server; the central server decrypts the encrypted information, specifically including: the central server also uses the designated user number to perform encryption calculation, generates a verification value, and compares it with the encrypted information sent by the user mobile terminal; wherein the designated user number is the initial user number, that is, the unique registration number generated when the user registers; the initial user number is only known to the user and the administrator of the central server, and the user can modify it through a specific algorithm; the mixed encryption algorithms used by the user mobile terminal include MD4, MD5, HASH and SM3 to achieve encryption and decryption authentication.

[0111] The calculation using the user number uses a digest algorithm, so after the user's mobile terminal sends encrypted information, the central server must also perform encryption calculations to obtain it. This algorithm is used for encryption authentication. The user's mobile terminal uses asymmetric encryption algorithms such as RSA and SM2 to achieve the transmission of keys used in symmetric encryption algorithms; the user's mobile terminal uses an asymmetric encryption algorithm to encrypt the key and sends the encrypted ciphertext to the central server; the central server decrypts the ciphertext to obtain the key.

[0112] Specifically, it also includes: setting up a safety alarm device to promptly report an emergency, such as a fire or vehicle breakdown, and start an emergency rescue mechanism to ensure the safety of vehicles and personnel.

[0113] Specifically, the optimal driving path model is constructed, and after identification, the optimal driving path is planned for the user, including:

[0114] Get the endpoint position of the vehicle and the endpoint position of the parking transition area,

[0115] Calculate the best ordered inflection point combination between the vehicle position and the parking transition area according to the relationship between the endpoint position of the vehicle and the endpoint position of the parking transition area;

[0116] The first endpoint information in the best ordered inflection point combination is extracted and sent to the user's mobile terminal; when the endpoint position of the vehicle changes, the optimal driving path is recalculated.

[0117] Get the endpoint location of the vehicle: Monitor the real-time location information of the vehicle or obtain the current location of the vehicle through the GPS system. Determine the endpoint location of the parking transition area based on the location information of the parking lot or parking space. Determine the best driving path based on the distance between the two and the road conditions. Send the endpoint information to the user's mobile terminal to guide the user's driving direction. Monitor changes in the vehicle's position. If there are any changes, recalculate the best driving path and update the navigation information on the user's mobile terminal.

[0118] Specifically, based on the improved Mask-RCNN network model, the vehicle is safely parked in a suitable parking space through mechanical equipment, including:

[0119] Collect images of the parking status of various vehicles in parking spaces, and divide them into sample data sets and data of vehicles to be parked;

[0120] Input the sample data set and the data of the vehicles to be parked into the CNN network of the improved Mask-RCNN network model to extract image features;

[0121] The feature map is obtained by integrating image features through the improved EC module; the ROI of the feature map is extracted and filtered through the RPN network;

[0122] The region of interest alignment layer is used to unify the size of the feature map according to the filtered ROI. Finally, the feature map is subjected to binary classification, bounding box regression, and mask segmentation in the fully connected layer to generate a binary mask image, and a multi-task loss function L of the binary mask image is established.

[0123] The value of the multi-task loss function L is reduced by continuously adjusting the parameters of binary classification, bounding box regression, and Mask until the global optimal solution of the multi-task loss function L is obtained;

[0124] The global optimal solution and weight matrix of the multi-task loss function L are substituted into the vehicle safe parking function, and the Mask-RCNN network model parameters are improved through multiple iterations until the optimal solution of the vehicle safe parking function is obtained; according to the optimal solution, the mechanical equipment is controlled to safely park the vehicle in an available parking space.

[0125] Specifically, the improved EC module includes a first channel and a second channel, the first channel is composed of a first convolution unit and a second convolution unit, the second channel is composed of a second convolution unit, the input image features enter the first convolution unit in the first channel and the second convolution unit in the second channel respectively, the output result of the first convolution unit in the first channel enters the second convolution unit in the first channel, and then the output results of the second convolution unit in the first channel and the second convolution unit in the second channel are feature connected by weighted summation, and finally the output layer outputs the integrated image features.

[0126] The first convolution unit and the second convolution unit use Leaky Relu activation function and Linear activation function respectively. A group self-adaptive normalization layer GSN is set after the first convolution unit and the second convolution unit to normalize the convolution result. The group self-adaptive normalization layer GSN consists of a batch normalization layer BN, an instance normalization layer IN, and a layer normalization layer GN.

[0127] The purpose of designing a multi-task loss function to find its global optimal solution is to achieve pixel-level image target segmentation. When the value of the multi-task loss function is the global optimal solution, the detection effect of the sample data set reaches the best. Considering the problem of similarity, this method combines the weight fusion similarity with the image features, calculates the weight matrix, and finally uses the weight matrix and the global optimal solution of the multi-task loss function to obtain the optimal solution of the image processing objective function, thereby improving the accuracy and precision of the recognition results. Therefore, the present invention can improve the accuracy and precision of the vehicle recognition results on the parking space, and the detection effect is good.

[0128] The improved EC module includes a first channel and a second channel, and the image features are input into the first channel and the second channel respectively, and then the output results of the first channel and the second channel are connected by weighted feature, and finally the output layer outputs the integrated image features, the first channel is composed of a first convolution unit and a second convolution unit, the second channel is composed of a second convolution unit, and the first convolution unit and the second convolution unit are both provided with a group self-adaptive normalization layer GSN for normalizing the convolution results, and the group self-adaptive normalization layer GSN is composed of a batch normalization layer BN, an instance normalization layer IN, and a layer normalization layer GN; in this design, the group self-adaptive normalization layer GSN not only enhances the adaptability of network feature information aggregation, image information robustness is good, and the accuracy of target detection tasks is improved, but also has low computational complexity, simplifies the computational redundancy of the feature processing process, and ultimately reduces the recognition time consumption while improving the recall rate. Therefore, the present invention reduces the recognition time consumption while improving the recall rate by improving the EC module.

[0129] Specifically, the weight matrix is ​​expressed as:

[0130] ω l =[d1, d2…d i …d n ];

[0131] n is the number of sample data sets and vehicle data to be parked,

[0132] d i The expression is as follows:

[0133]

[0134] In the above formula, α C , α S , α H are the actual contributions of the color feature vector, texture feature vector, and shape feature vector of the parking state of the vehicle in the parking space to the similarity measurement; are the actual contributions of the total similarity measure of the N1 color feature vector, N2 texture feature vector, and N3 shape feature of the parking state of the vehicle in the parking space, are the color feature vector, texture feature vector, and shape feature vector of the i-th image of the vehicle to be parked in the parking space, respectively. are respectively the color feature vector, texture feature vector, and shape feature vector of the i-th image in the sample data set of the vehicle to be parked in the parking state in the parking space, P is the vehicle data to be parked, T is the sample data set, and i is the time sequence number of the image in the data set.

[0135] Specifically, the expression of the vehicle safe parking function f(x) is as follows:

[0136]

[0137] In the formula, L is the loss function, H l is the activation function, x i is the vehicle to be parked, ω l is the weight matrix, y i is the grayscale image function, and n is the number of sample data sets and vehicle data to be parked.

[0138] The optimal solution of the vehicle safety parking function f(x) is expressed as follows:

[0139]

[0140] In the above formula, x i is a sample data set, H l ' is the activation function H l The partial derivative of the weight matrix, η is the regularization coefficient.

[0141] Specifically, a path planning algorithm is used to calculate the optimal driving path, including:

[0142] Determine the endpoint position of the vehicle, the endpoint position of the parking transition area, and the time when the user arrives at the garage;

[0143] Obtain all available driving paths based on the endpoint position of the vehicle and the endpoint position of the parking transition area; calculate the distance, time and congested vehicles ahead of each driving path;

[0144] An initial planning path is obtained from all driving paths according to the saving algorithm; the inflection point closest to the vehicle's endpoint is selected as the first inflection point, the estimated time from other inflection points to the next inflection point and the estimated time from the first inflection point to the next inflection point are calculated, and the time difference is evaluated according to the saving algorithm to determine whether to use other inflection points as selected inflection points until the iteration condition is met, and all selected inflection points are arranged in chronological order to form an initial planning path;

[0145] A large number of initial populations are generated from all the initial planning paths obtained, and the optimal driving path is obtained by iteration combined with genetic algorithm.

[0146] Specifically, all the initial planned paths are arranged in random order, each arrangement is used as a chromosome, and a plurality of the chromosomes constitute the initial population.

[0147] Genetic Algorithm is a computational model of biological evolution that simulates the natural selection and genetic mechanisms of Darwin's theory of biological evolution. It is a method of searching for the optimal solution by simulating the natural evolution process.

[0148] The embodiment of the present invention arranges all the initial planning paths obtained in random order, and regards each arrangement as a chromosome. A plurality of the chromosomes constitute an initial population, and the optimal planning path is obtained by iteration in combination with a genetic algorithm, wherein the chromosome composition is a complete gene individual generated by encoding the coordinates of each inflection point in the initial planning path in sequence, and encoding the paths with special symbols. At the same time, the chromosome is copied and expanded to form a population, and each chromosome can obtain the fitness value of the chromosome, that is, the shortest time value, through the fitness evaluation function. The next generation of chromosomes is obtained by crossing any two chromosomes in the initial population. If the fitness value of the next generation of chromosomes is less than that of the parent generation, it is retained, and the one with the smaller value in the parent generation is copied, and it is inherited from generation to generation until an individual with the most suitable fitness value is obtained as the optimal driving path.

[0149] Specifically, the saving algorithm and the genetic algorithm implement the operation process under the spark architecture based on the Hadoop distributed cluster.

[0150] Specifically, the genetic algorithm includes obtaining an initialized population, performing fitness detection on the initialized population, performing selection, crossover and mutation operations on the population after fitness monitoring, and judging whether it meets the termination condition. If the termination condition is not met, re-performing fitness detection, and repeating the selection, crossover and mutation operations, if the termination condition is met, performing the termination operation, the termination condition is a preset number of iterations, and the fitness detection refers to retaining the solution with the larger fitness evaluation function, and the fitness evaluation function is expressed as:

[0151]

[0152] Among them, f i (T) is the fitness value of chromosome i in the Tth generation population, z i (T) is the objective function value of chromosome i in the T-generation population, that is, the total transportation cost corresponding to chromosome i, f i The larger the (T) value is, the closer the route corresponding to chromosome i is to the optimal driving path.

[0153] The embodiments of the present invention can intelligently implement path planning and find the optimal driving path for users.

[0154] In a specific embodiment of the present invention, a highly secure mechanical stereo garage intelligent parking system, such as Figure 2 As shown, including:

[0155] The information receiving module is used to receive parking information or pick-up information provided by the parking user through the garage central server; the parking information includes the user's license plate number, parking arrival time and estimated parking duration information; the pick-up information includes the user's pick-up arrival time;

[0156] The vehicle identification module is used to identify the vehicle according to the license plate number information when the vehicle arrives at the designated location;

[0157] The path planning module is used to build the optimal driving path model. After identification, it plans the optimal driving path for the user and guides the user into the parking transition area;

[0158] A weight information acquisition module is used to bury a weighing sensor under the parking transition area. After a vehicle enters the parking transition area, the weighing sensor acquires the weight information of the vehicle and determines the load-bearing capacity of each parking space according to the weight information of the vehicle;

[0159] A size information acquisition module is used to install an infrared sensor at the entrance or around the parking transition area to acquire the size information of the vehicle;

[0160] A vehicle parking module, which is used to intelligently evaluate the weight and size information of the vehicle, determine an available parking space suitable for the vehicle, and safely park the vehicle in the available parking space through mechanical equipment;

[0161] The vehicle extraction module is used to move the vehicle from the parking space to the parking transition area through mechanical equipment according to the vehicle extraction information.

[0162] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0163] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A highly secure mechanical garage intelligent parking method, characterized in that: include: The parking information or pick-up information provided by the parking user is received through the garage central server; the parking information includes the user's license plate number, parking arrival time and estimated parking duration information; the pick-up information includes the user's pick-up arrival time; When the vehicle arrives at the designated location, the vehicle is identified based on the license plate number information; Build an optimal driving path model, and after identification, plan the optimal driving path for the user and guide the user into the parking transition area; A weighing sensor is buried under the parking transition area. After a vehicle enters the parking transition area, the weighing sensor obtains the weight information of the vehicle, and the load-bearing capacity of each parking space is determined according to the weight information of the vehicle; Install infrared sensors at the entrance or around the parking transition area to obtain the size information of the vehicle; Intelligently evaluate the weight and size information of the vehicle, determine the available parking space suitable for the vehicle, and park the vehicle safely in the available parking space through mechanical equipment; According to the vehicle pick-up information, the vehicle is moved from the parking space to the parking transition area by mechanical equipment.

2. A highly secure mechanical garage intelligent parking method according to claim 1, characterized in that: Also includes: Encryption technology is used during data transmission and storage to encrypt and protect the information and parking records provided by users.

3. A highly secure mechanical garage intelligent parking method and system according to claim 1, characterized in that: Intelligently evaluate the vehicle's weight and size information to determine available parking spaces suitable for the vehicle; Park the vehicle safely in an available parking space using mechanical equipment, including: Set a maximum load-bearing threshold for each parking space, analyze the weight data of each parking space based on an intelligent algorithm, and automatically mark a vehicle as unavailable when its weight reaches or exceeds the threshold; The intelligent algorithm analyzes the maximum load-bearing capacity of each parking space based on the current vehicle weight data, and evaluates the stress condition of the entire mechanical parking garage in combination with the weight data of all parking spaces; Set the maximum size threshold for each parking space, compare and analyze the size information of each parking space, and automatically mark a vehicle as unavailable when its size reaches or exceeds the threshold. When the vehicle's weight information and size information both meet the above conditions, it is determined to be an available parking space; Move the vehicle to an available parking space.

4. A highly secure mechanical garage intelligent parking method according to claim 3, characterized in that: Also includes: Install a weighing sensor on each parking space to obtain real-time weight data of each parking space; Update the usage status of each parking space in real time.

5. A highly secure mechanical garage intelligent parking method according to claim 3, characterized in that: The intelligent algorithm analyzes the maximum load-bearing capacity of each parking space based on the current vehicle weight data, and evaluates the stress conditions of the entire mechanical parking garage in combination with the weight data of all parking spaces, including: Establish a three-dimensional coordinate system for the garage, where (x, y, z) represents the actual coordinate center position of the parking space; Create parking space matrix R i (x, y, z), i represents the number of the parking space, x represents the x-axis coordinate of the i-th parking space, y represents the y-axis coordinate of the i-th parking space, and z represents the z-axis coordinate of the i-th parking space; Initialize n smart matching operators. The smart matching operators work in a cyclic calculation mode. Set the total number of cyclic calculations to c. Each smart matching operator can perform matching calculations and position change calculation tasks. Establish the smart matching operator matrix M. k (x'k(j), y'k(j), z'k(j), P), k represents the number of the smart matching operator, x'k(j) represents the x-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, y'k(j) represents the y-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, z'k(j) represents the z-axis coordinate position of the k-th smart matching operator during the j-th loop calculation, and P represents the vehicle information loaded by the k-th smart matching operator; the information of available parking spaces in the garage is loaded into the smart matching operator matrix, and the random number generation algorithm is used to randomly assign x, y, and z coordinates to each smart matching operator, thereby initializing the smart matching operator to the three-dimensional coordinate system of the garage; The initialized smart matching operator performs a matching calculation if there is an available parking space at the current initialization position. If the matching result meets the maximum load-bearing and maximum accommodating size thresholds of the parking space, it is considered a successful match and the matching result Sui is recorded in the matching history information table His; If the matching result does not meet the maximum load-bearing and / or maximum accommodating size thresholds of the parking space, the matching is unsuccessful, and the position information will be calculated using the smart matching operator displacement algorithm based on the conversion switch in the next cycle of the smart matching operator.

6. A highly secure mechanical garage intelligent parking method according to claim 1, characterized in that: Build an optimal driving path model. After identification, plan the optimal driving path for the user, including: Get the endpoint position of the vehicle and the endpoint position of the parking transition area, Calculate the best ordered inflection point combination between the vehicle position and the parking transition area according to the relationship between the endpoint position of the vehicle and the endpoint position of the parking transition area; The first endpoint information in the best ordered inflection point combination is extracted and sent to the user's mobile terminal; when the endpoint position of the vehicle changes, the optimal driving path is recalculated.

7. A highly secure mechanical garage intelligent parking method according to claim 1, characterized in that: Based on the improved Mask-RCNN network model, the vehicle is safely parked in a suitable parking space through mechanical equipment, including: Collect images of the parking status of various vehicles in parking spaces, and divide them into sample data sets and data of vehicles to be parked; Input the sample data set and the data of the vehicles to be parked into the CNN network of the improved Mask-RCNN network model to extract image features; The feature map is obtained by integrating image features through the improved EC module; the ROI of the feature map is extracted and filtered through the RPN network; The region of interest alignment layer is used to unify the size of the feature map according to the filtered ROI. Finally, the feature map is subjected to binary classification, bounding box regression, and mask segmentation in the fully connected layer to generate a binary mask image, and a multi-task loss function L of the binary mask image is established. The value of the multi-task loss function L is reduced by continuously adjusting the parameters of binary classification, bounding box regression, and Mask until the global optimal solution of the multi-task loss function L is obtained; The global optimal solution and weight matrix of the multi-task loss function L are substituted into the vehicle safe parking function, and the Mask-RCNN network model parameters are improved through multiple iterations until the optimal solution of the vehicle safe parking function is obtained; according to the optimal solution, the mechanical equipment is controlled to safely park the vehicle in an available parking space.

8. A highly secure mechanical garage intelligent parking method according to claim 7, characterized in that: The expression of the vehicle safe parking function f(x) is as follows: In the formula, L is the loss function, H l is the activation function, x i is the vehicle to be parked, ω l is the weight matrix, y i is the grayscale image function, n is the number of sample data sets and vehicle data to be parked; The optimal solution of the vehicle safety parking function f(x) is expressed as follows: In the above formula, x i is a sample data set, H l ' is the activation function H l The partial derivative of the weight matrix, η is the regularization coefficient.

9. A highly secure mechanical garage intelligent parking method according to claim 6, characterized in that: Use path planning algorithms to calculate the optimal driving path, including: Determine the endpoint position of the vehicle, the endpoint position of the parking transition area, and the time when the user arrives at the garage; Obtain all selectable driving paths according to the endpoint position of the vehicle and the endpoint position of the parking transition area; An initial planning path is obtained from all driving paths according to the saving algorithm; the inflection point closest to the vehicle's endpoint is selected as the first inflection point, the estimated time from other inflection points to the next inflection point and the estimated time from the first inflection point to the next inflection point are calculated, and the time difference is evaluated according to the saving algorithm to determine whether to use other inflection points as selected inflection points until the iteration condition is met, and all selected inflection points are arranged in chronological order to form an initial planning path; A large number of initial populations are generated from all the initial planned paths obtained, and the optimal driving path is obtained by iteration combined with genetic algorithm.

10. A highly secure mechanical stereo garage intelligent parking system, characterized in that: include: The information receiving module is used to receive parking information or pick-up information provided by the parking user through the garage central server; the parking information includes the user's license plate number, parking arrival time and estimated parking duration information; the pick-up information includes the user's pick-up arrival time; The vehicle identification module is used to identify the vehicle according to the license plate number information when the vehicle arrives at the designated location; The path planning module is used to build the optimal driving path model. After identification, it plans the optimal driving path for the user and guides the user into the parking transition area; A weight information acquisition module is used to bury a weighing sensor under the parking transition area. After a vehicle enters the parking transition area, the weighing sensor acquires the weight information of the vehicle and determines the load-bearing capacity of each parking space according to the weight information of the vehicle; A size information acquisition module is used to install an infrared sensor at the entrance or around the parking transition area to acquire the size information of the vehicle; A vehicle parking module, which is used to intelligently evaluate the weight and size information of the vehicle, determine an available parking space suitable for the vehicle, and safely park the vehicle in the available parking space through mechanical equipment; The vehicle extraction module is used to move the vehicle from the parking space to the parking transition area through mechanical equipment according to the vehicle extraction information.

Citation Information

Cited By

  • Three-dimensional intelligent scheduling platform and scheduling method under multi-mode traffic fusion

    CN119851503A

  • Three-dimensional intelligent dispatching platform and method under multi-modal traffic integration

    CN119851503B