AI multi-currency change parking control system

Abstract

The invention, which relates to the technical field of artificial intelligence, discloses an AI multi-currency change parking control system comprising a central calculation module, a currency identification module, a classified access module and an interactive display module. The central calculation module accumulates the parking time length based on the vehicle license plate and the confirmation signal, sends the total consumption amount, calculates the total payment amount and the total change amount through an accumulative change decision method, and generates a change scheme; the currency identification module obtains a payment currency grey-scale map and cooperatively identifies payment real currency and payment amount by adopting KL transformation and a genetic evolution back propagation network; the classified access module stores payment genuine coins in a classified manner, executes a change scheme, and decides whether to send a change stock supplement prompt or a full stock prompt based on a stock change principle; and the interactive display module senses user operation to generate a vehicle license plate or a confirmation signal, and displays the display information in real time, so that an automatic intelligent system integrating three functions of accurate recognition, optimal change and automatic supplementary prompt is realized.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an AI multi-currency change parking control system. Background Art

[0002] In traditional parking management systems, the fee collection and change process for vehicles entering and exiting parking lots relies primarily on manual operations or simple automated equipment. Manual fee collection is inefficient, prone to traffic congestion during peak hours, and has high labor costs. It also carries the risk of human error, such as incorrect change.

[0003] Existing automated parking meter systems typically only support a single currency or a few common payment methods. This makes them unable to meet the diverse payment needs of users from diverse regions in highly internationalized areas or tourist attractions. Furthermore, these devices can suffer from insufficient accuracy and slow response times when processing complex multi-currency change calculations, making it difficult to achieve fast and accurate change, impacting parking efficiency and user experience. Therefore, currency recognition and automatic change are key components of automated parking systems. Summary of the Invention

[0004] In response to the shortcomings of the existing technology, the present invention proposes an AI multi-currency change parking control system to accurately identify the authenticity and face value of currency, and provide change solutions and supplementary prompt functions.

[0005] The technical solutions for achieving the purpose of the present invention are:

[0006] An AI multi-currency change parking control system includes a central computing module, a currency recognition module, a classification access module, an interactive display module, a currency transaction entrance, and a currency transmission channel;

[0007] The central calculation module confirms the vehicle information and parking time based on the vehicle license plate, and accumulates the parking time to calculate the total consumption amount x b , based on the confirmation signal, generate and send the total consumption amount x b Go to the interactive display module and generate the total payment amount x by using the cumulative change decision method p and the total amount of change x r Send to the interactive display module to generate a change plan And send it to the classification access module;

[0008] The currency recognition module obtains the grayscale image of the payment currency and uses KL transformation to perform dimensionality reduction and feature extraction to generate effective features. It uses the genetic evolution back propagation network to analyze the effective features to identify the payment of real currency or counterfeit currency and deliver it to the currency transmission channel, sending the payment amount and face value mark to the central calculation module and classification access module respectively;

[0009] The classification access module stores payment real coins based on face value tags, receives and executes the change solution After the change is completed, a recovery signal is fed back to the interactive display module, and based on the inventory change principle, a statistical decision is made whether to send a change inventory replenishment prompt or a full inventory prompt;

[0010] The interactive display module senses the user's operation to generate a vehicle license plate or confirmation signal and sends it to the central computing module, which receives the display information and displays it in real time. When it receives the recovery signal, it clears the display information. The display information includes vehicle information, parking time, total consumption amount x b 、Total payment amount x p and the total amount of change x r ;

[0011] The currency transaction port is used to receive payment currency, return counterfeit payment currency, and return change currency;

[0012] The currency transmission channel is used to transmit payment currency, payment of counterfeit currency, payment of real currency and change currency.

[0013] Furthermore, the central computing module includes a main memory unit, a statistics unit, and a change computing unit;

[0014] The main memory unit stores a parking time table, and the vehicle information and parking time corresponding to the vehicle license plate can be obtained by querying the parking time table;

[0015] The statistics unit queries the parking time table based on the vehicle license plate to obtain the corresponding vehicle information and parking time and sends it to the interactive display module. It updates the total consumption amount to the current total consumption amount plus the parking time, receives the confirmation signal and sends the total consumption amount x b Go to the interactive display unit, open the currency transaction portal and start the currency identification module and currency transmission channel;

[0016] The change calculation unit receives the payment amount and sends the total payment amount x to the interactive display module based on the cumulative change decision method. p and the total amount of change x r , generate change plan And send it to the classification access module to clear the total consumption amount x b and the total payment amount x p , close the currency transaction entrance and stop the currency identification module and currency transmission channel.

[0017] Furthermore, the cumulative change decision-making method includes the following specific steps:

[0018] Get the total consumption amount x b and the change currency inventory quantity sequence n x , set the total payment amount x p is 0;

[0019] Receive payment amount and update total payment amount x p The current total payment amount plus the payment amount;

[0020] Determine the total payment amount x p Is it greater than or equal to the total consumption amount x b ;

[0021] If it is less than the total consumption amount x b , waiting to receive new payment amount;

[0022] If it is greater than or equal to the total consumption amount x b , calculate the total change amount x r and with the total payment amount x p Send it to the interactive display module together;

[0023] Construct the change optimization target as the total number of used Minimum, set the change optimization constraint 1 to the total amount of change currency used equal to the total amount of change x r , set the change optimization constraint 2 to: the number of change coins of a single denomination used must be less than or equal to the corresponding change coin inventory quantity;

[0024] Combine the change optimization objective, change optimization constraint 1, and change optimization constraint 2 to generate the change optimization problem, and use the improved greedy algorithm to solve and obtain the change solution. And sent to the classification storage module.

[0025] Furthermore, the improved greedy algorithm solves the change optimization problem to obtain the change solution The specific steps include:

[0026] Initialize the remaining total amount and change plan The total amount of change is x r and all-0 vector;

[0027] Execute change currency selection, use the change currency in stock corresponding to the largest currency denomination that is less than or equal to the current remaining total amount as the change currency, and determine whether the product of the change currency denomination and the change currency stock quantity is greater than the current remaining total amount;

[0028] If it is greater than the current remaining total amount, the maximum number of change currency usages that satisfies the requirement that the face value of the change currency multiplied by the number of change currency usages is less than or equal to the current remaining total amount will be used as the optimal number of change currency usages;

[0029] If it is less than or equal to the current remaining total amount, the optimal amount of change currency used is equal to the amount of change currency in stock;

[0030] Put the optimal change currency usage into the change plan The corresponding position in

[0031] Update the remaining total amount to the current remaining total amount minus the product of the optimal change currency usage and the change currency denomination, and determine whether the remaining total amount is equal to 0;

[0032] If the remaining total amount is not equal to 0, jump to execute the change currency selection. If the remaining total amount is equal to 0, get the change solution

[0033] Furthermore, the currency recognition module includes a scanning device, a feature extraction unit, and an identification marking unit;

[0034] The scanning module is used to obtain the grayscale image of the payment currency;

[0035] The feature extraction module uses KL transformation to calculate the covariance matrix of the payment currency grayscale image and perform eigenvalue decomposition. The contribution of the corresponding eigenvector is determined based on the eigenvalue size, and multiple eigenvectors with the largest contribution are selected to generate effective features.

[0036] The identification marking unit establishes a nonlinear mapping from effective features to the authenticity and amount of currency through a genetic evolution back propagation network. If it is a payment of real currency, the payment amount and face value mark are generated and sent to the central computing module and the classification access module respectively. The real currency is delivered to the currency transmission channel. If it is a payment of counterfeit currency, it is delivered to the currency transmission channel.

[0037] Furthermore, the training of the genetic evolution back propagation network includes the following specific steps:

[0038] The initial population is randomly generated, and each individual in the initial population is processed using floating-point encoding to generate chromosomes in the genetic evolution process;

[0039] The fitness function is set as the inverse of the sum of squares of the error between the output value of the back propagation network and the actual value, which is used to influence the inheritance probability in the genetic evolution process;

[0040] The reproduction operator is determined to be the fitness ratio method, that is, the selection probability of a single individual is the fitness function value of a single individual divided by the sum of the fitness function values ​​of all individuals;

[0041] The crossover operator is determined to be partial arithmetic crossover, which is used to extract partial chromosomes from the chromosomes of the two individuals, perform linear combination, and then fill the vacant parts of the original chromosomes to generate new individuals;

[0042] Determine the mutation operator of the genetic evolution process. The mutation probability of the mutation operator and the evolutionary generation are normally distributed, which plays a fine-tuning role.

[0043] Set the maximum evolutionary generation, calculate the fitness function values ​​of all individuals in the initial population, and determine whether the maximum evolutionary generation is met;

[0044] If the maximum evolutionary generation is not satisfied, the two individuals with the largest reproduction operator are selected, and new individuals are generated based on the crossover operator and mutation operator.

[0045] If the maximum evolutionary generation is met, the chromosome of the individual with the largest fitness function value in this generation of genetic evolution is retained as the threshold and connection weight of the back propagation network;

[0046] Perform forward propagation of the backpropagation network, calculate the loss function value, and determine whether the loss function value is less than or equal to the threshold;

[0047] If it is greater than the threshold, the gradient descent method is used to adjust the connection weights, recalculate the loss function value and continue to judge whether it is less than or equal to the threshold. If it is less than or equal to the threshold, the training is completed.

[0048] Furthermore, the classified access module includes a cash box and an access control unit;

[0049] The cash box is used to store currencies of different denominations in a classified manner;

[0050] The access control unit controls the classified storage of payment real coins based on the face value mark, and based on the change scheme The change currency is extracted in sequence and delivered to the currency transmission channel. After the change is completed, the recovery signal is fed back to the interactive display module. Based on the inventory change principle, a statistical decision is made whether to send a change inventory replenishment prompt or a full inventory prompt.

[0051] Furthermore, the access control unit statistically determines whether to send a change inventory replenishment prompt or a non-change full inventory prompt based on the inventory change principle, including:

[0052] When the real currency is deposited into the cash box, the change currency inventory quantity sequence n x The amount of currency in stock corresponding to the face value of the real currency to be paid is increased by 1, and the amount of currency in stock corresponding to the face value of the currency is judged to see whether it has reached the storage limit. If so, a full storage prompt is sent; if not, no action is taken.

[0053] When the change is completed, re-count and generate the change currency inventory quantity sequence n x , synchronously execute the first proposition judgment and the second proposition judgment to determine whether the change inventory currency can meet all the amounts in 1-99. If it does, no processing is done. If it does not, a change inventory replenishment reminder is sent.

[0054] Furthermore, the interactive display module includes a touch display screen, an induction encoding unit and a display control unit;

[0055] The touch display screen receives the display image and updates the display based on the increase of local charge caused by the user's touch;

[0056] The sensing encoding unit obtains the vehicle license plate, decides whether to generate a confirmation signal based on the sensing effectiveness strategy, and sends the vehicle license plate or confirmation signal to the central computing module;

[0057] The display control unit adds or replaces the display information to the current rendering template when it receives the display information, clears the display information in the current rendering template when it receives the restoration signal, renders and generates a display image and sends it to the touch screen.

[0058] Furthermore, the sensing encoding unit decides whether to generate a confirmation signal based on the sensing validity strategy, including the following specific steps:

[0059] Detecting the increase of local charge, obtaining the local charge peak value and determining whether it is greater than the effective touch valley value;

[0060] If it is less than or equal to the effective touch valley value, ignore the increase in local charge;

[0061] If it is greater than the effective touch valley value, the range of the local charge increase area is confirmed, compared with the range of the confirmation button and the overlap ratio is calculated;

[0062] If the coincidence ratio is greater than the coincidence ratio threshold, a confirmation signal is generated; if the coincidence ratio is less than or equal to the coincidence ratio threshold, the local charge increase is ignored.

[0063] Compared with the prior art, the present invention obtains payment currency based on multiple scanning methods, and collaboratively identifies real payment currency and payment amount in a genetic evolution back propagation network with optimal structure and connection weights optimized by KL transformation and improved greedy algorithm, thereby improving the accuracy of currency recognition. It adopts a cumulative change decision method to calculate the total payment amount and the total change amount and generate a change plan that takes into account the amount of inventory currency and minimizes the total number of change currencies, thereby reducing the frequency of currency replenishment and improving the user's change experience. It decides whether to send a change inventory replenishment prompt or a full inventory prompt based on the inventory change principle to promptly respond to the problems of insufficient and full inventory currency. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A schematic diagram of a robotic parking device used in a parking control system;

[0065] Figure 2 This is a rear view of the robotic parking device;

[0066] Figure 3 This is a schematic diagram of an AI multi-currency change parking control system;

[0067] Figure 4 This is a flow chart of the cumulative change decision method in the present invention;

[0068] Figure 5 This is a flow chart of the improved greedy algorithm in the present invention;

[0069] Figure 6 This is a flow chart of the effective sensing strategy in the present invention. DETAILED DESCRIPTION

[0070] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0071] Example 1

[0072] like Figures 1 to 3 As shown, a specific embodiment of the present invention discloses an AI multi-currency change parking control system, including a robotic parking device using the AI ​​multi-currency change parking control system to facilitate the display of parking-related information and currency insertion and change. The AI ​​multi-currency change parking control system includes a central computing module, a currency recognition module, a classification access module, an interactive display module, a currency transaction entrance, and a currency transmission channel.

[0073] The central calculation module determines the vehicle information and parking time based on the vehicle license plate and accumulates them to calculate the total consumption amount x b , based on the confirmation signal, send the total consumption amount x b Go to the interactive display module, obtain the payment amount, and calculate the total payment amount x using the cumulative change decision method p and the total amount of change x r And send it to the interactive display module to generate a change plan And send it to the classification access module;

[0074] The currency recognition module obtains a grayscale image of the payment currency through red light, infrared light, and fluorescent light scanning, and uses KL transformation to reduce the dimension and extract features of the second-order statistical information in the payment currency grayscale image to generate effective features. A nonlinear mapping from effective features to currency authenticity and currency amount is established through a genetic evolution back propagation network. The module identifies whether the payment is genuine or counterfeit currency and delivers it to the currency transmission channel. If it is identified as a genuine payment, the currency denomination corresponding to the genuine payment is sent to the central computing module as the payment amount, and an additional denomination tag is generated and sent to the classification access module.

[0075] The classification access module stores the payment real currency based on the face value mark, receives and executes the change plan The change currency is delivered to the currency transmission channel, and after the change is completed, a recovery signal is fed back to the interactive display module. Based on the inventory change principle, a statistical decision is made whether to send a change inventory replenishment prompt or a full inventory prompt.

[0076] The interactive display module senses the user's operation to generate a vehicle license plate or confirmation signal and sends it to the central computing module, which receives the display information and displays it in real time on the touch screen. When it receives the recovery signal, it clears the display information. The displayed information includes vehicle information, parking time, total consumption amount x b 、Total payment amount x p and the total amount of change x r ;

[0077] The currency transaction port includes a currency entry and a currency exit. The currency entry is used to receive payment currency and return counterfeit payment currency, and the currency exit is used to return change currency.

[0078] The currency transmission channel includes a bidirectional payment channel, a real currency storage channel, and a change channel. The bidirectional payment channel connects the currency entrance and the currency identification module, and is used to transmit the payment currency to the currency identification module and return the payment counterfeit currency to the currency entrance. The real currency storage channel connects the currency identification module and the classification access module, and is used to transmit the payment real currency to the classification access module. The change channel connects the classification access module and the currency exit, and is used to transmit the change currency to the currency exit.

[0079] Furthermore, the central computing module includes a main memory unit, a statistics unit, and a change computing unit;

[0080] The main memory unit stores a parking time table, which is used to record vehicle information and parking time corresponding to the vehicle license plate;

[0081] The statistics unit receives the vehicle license plate, queries the parking time table to obtain the corresponding vehicle information and parking time, obtains the current total consumption amount, updates the total consumption amount to the current total consumption amount plus the parking time, sends the vehicle information and parking time to the interactive display module, receives the confirmation signal, and sends the total consumption amount x b Go to the interactive display unit, open the currency transaction portal and start the currency identification module and currency transmission channel;

[0082] The change calculation unit receives the payment amount and sends the total payment amount x to the interactive display module based on the cumulative change decision method. p and the total amount of change x r , generate change plan And send it to the classification access module to clear the total consumption amount x b and the total payment amount x p , close the currency transaction entrance and stop the currency identification module and currency transmission channel.

[0083] like Figure 4 As shown, further, the cumulative change decision method includes the following specific steps:

[0084] Get the total consumption amount x b and the change currency inventory quantity sequence nx =[n 50 ,n 20 ,n 10 ,n5,n1], set the total payment amount x p is 0, where n i Indicates the stock quantity of change currency with a denomination of i;

[0085] Receive the payment amount, let the total payment amount be x p Updated to the current total payment amount plus the payment amount;

[0086] Determine the total payment amount x p Is it greater than or equal to the total consumption amount x b ;

[0087] If the total payment amount is x p Less than the total amount of consumption x b , waiting to receive new payment amount;

[0088] If the total payment amount is x p Greater than or equal to the total consumption amount x b , calculate the total change amount x r =x p -x b , send payment total amount x p and the total amount of change x r To the interactive display module;

[0089] Construct the change optimization target, the change optimization target is the total number of uses Minimum, i.e. min Total number of uses It is the sum of the number of change currencies used for each currency denomination. The specific formula is as follows:

[0090]

[0091] Among them, x = {50, 20, 10, 5, 1} is the set of change currency denominations, is the number of change coins used with a face value of i;

[0092] Set change optimization constraint 1, which requires that the total amount of change currency used is equal to the total change amount x r ,Right now

[0093] Set the change optimization constraint 2, which requires that the number of change coins of a single denomination used must be less than or equal to the corresponding change coin inventory quantity, that is, i∈x;

[0094] Combine the change optimization objective, change optimization constraint 1, and change optimization constraint 2 to generate a change optimization problem. Use the improved greedy algorithm to solve the change optimization problem and obtain the change solution. And sent to the classification storage module.

[0095] like Figure 5 As shown, further, the improved greedy algorithm solves the change optimization problem to obtain the change solution The specific steps include:

[0096] Initialize the remaining total amount to the total change amount x r , set up the change plan is a full 0 vector;

[0097] Execute change currency selection, use the change currency in stock corresponding to the largest currency denomination that is less than or equal to the current remaining total amount as the change currency, and obtain the change currency denomination and change currency stock quantity of the change currency;

[0098] Determine whether the product of the change currency denomination and the change currency inventory quantity is greater than the current remaining total amount;

[0099] If it is greater than the current remaining total amount, the maximum number of change currency usages that satisfies the requirement that the face value of the change currency multiplied by the number of change currency usages is less than or equal to the current remaining total amount will be used as the optimal number of change currency usages;

[0100] If it is less than or equal to the current remaining total amount, all change currency stocks are selected, and the optimal number of change currencies to use is equal to the number of change currency stocks;

[0101] Place the optimal number of change currencies into the change plan The position with the same denomination as the change currency;

[0102] Update the remaining total amount to the current remaining total amount minus the product of the optimal change currency usage and the change currency denomination, and determine whether the remaining total amount is equal to 0;

[0103] If the remaining total amount is not equal to 0, jump to the change currency selection. If the remaining total amount is equal to 0, get the change solution

[0104] Furthermore, the currency recognition module includes a scanning device, a feature extraction unit, and an identification marking unit;

[0105] The scanning module acquires payment currency data information through various acquisition methods such as red light, infrared light and fluorescent scanning and saves it as a payment currency grayscale image with the suffix BDB;

[0106] The feature extraction module uses KL transformation to calculate the covariance matrix of the payment currency grayscale image and perform eigenvalue decomposition. The contribution of the eigenvector corresponding to the eigenvalue is determined according to the order of eigenvalues ​​from large to small. The largest eigenvector with a cumulative contribution rate greater than or equal to 90% is retained and combined to generate valid features.

[0107] The identification marking unit establishes a nonlinear mapping from effective features to the authenticity and amount of currency through a genetic evolution back propagation network. If the payment is real currency, the payment amount and face value mark are generated according to the face value of the real currency, the payment amount is sent to the central computing module, the real currency is delivered to the real currency storage channel in the currency transmission channel and the face value mark is synchronously sent to the classification access unit. If the payment is counterfeit currency, it is directly delivered to the two-way payment channel in the currency transmission channel to fall back to the entrance of the currency transaction port.

[0108] Furthermore, the genetic evolution back propagation network is improved based on the existing back propagation network. Since the back propagation network can ensure the final convergence of the network learning process, there is a problem of falling into the local optimal solution. At the same time, the performance of the back propagation network is closely related to the network structure, and the traditional network structure, including the number of network layers and the number of hidden layer nodes, are predetermined based on artificial experience. This means that even if the connection weights between the nodes of the back propagation network are adjusted through training to achieve the global optimal solution, it may not be the true global optimal solution. It is only the global optimal solution under this specific network structure. The genetic evolution algorithm has global optimization capabilities. By introducing the genetic evolution algorithm, the network structure and connection weights of the back propagation network are simultaneously optimized during the training phase to establish a genetic evolution back propagation network.

[0109] Furthermore, the training of the genetic evolution back propagation network includes the following specific steps:

[0110] Randomly generate an initial population, which includes multiple initial individuals. Each initial individual includes a threshold and multiple connection weights of different numbers and sizes. The number of connection weights essentially reflects the number of nodes in the back-propagation network, and the threshold provides a basis for judging the convergence of the back-propagation network loss function.

[0111] Floating-point encoding is used to process each individual in the initial population to generate chromosomes in the genetic evolution process. In this embodiment, the connection weights of each individual are cascaded in descending order, and a threshold is added at the end to generate a real array to implement floating-point encoding.

[0112] Set the fitness function, which is defined as the inverse of the sum of squares of the errors between the output value of the back propagation network and the actual value. The fitness function value is used to influence the inheritance probability in the genetic evolution process. Individuals with high fitness function values ​​have a greater probability of inheriting chromosomes to the next generation.

[0113] Determine the reproduction operator in the genetic evolution process. Since chromosomes are encoded as floating-point numbers, the reproduction operator uses the fitness ratio method, that is, the selection probability of a single individual is the fitness function value of the single individual divided by the sum of the fitness function values ​​of all individuals.

[0114] Determine a crossover operator in the genetic evolution process. The crossover operator is used to select two individuals from a population and transform parts of the chromosomes of the two individuals. This allows the genetic evolution process to be not limited to a randomly generated initial population and to explore more new chromosomes. In this embodiment, the crossover operator selects partial arithmetic crossover, which extracts parts of the chromosomes of the two individuals, performs linear combination, and then fills the gaps left by the original chromosomes to generate a new individual.

[0115] Determine the mutation operator of the genetic evolution process. The mutation operator is directly related to the evolutionary generation and plays a fine-tuning role. In the early stages of genetic evolution, the mutation range is large, allowing genetic evolution to explore more chromosomes. In the later stages of genetic evolution, the mutation range gradually narrows, approaching the optimal individual. In this embodiment, non-uniform mutation is selected, that is, the mutation rate of a single individual presents a normal distribution with the evolutionary generation. The larger the evolutionary generation, the closer the mutation rate is to 0.

[0116] Set the maximum evolutionary generation, calculate the fitness function value of each individual in the initial population, and determine whether the maximum evolutionary generation is met;

[0117] If the maximum evolutionary generation is not satisfied, the two individuals with the largest reproduction operator are selected, and new individuals are generated based on the crossover operator and mutation operator.

[0118] If the maximum evolutionary generation is met, the chromosome of the individual with the largest fitness function value in this generation of genetic evolution is retained as the threshold and connection weight of the back propagation network;

[0119] Perform forward propagation of the backpropagation network, calculate the loss function value, and determine whether the loss function value is less than or equal to the threshold;

[0120] If it is greater than the threshold, the gradient descent method is used to further adjust the connection weights, recalculate the loss function value and continue to judge whether it is less than or equal to the threshold. If it is less than or equal to the threshold, the back propagation network training is completed.

[0121] Furthermore, the classified access module includes a cash box and an access control unit;

[0122] The cash box includes a change storage box and a non-change storage box. The change storage box stores currencies with a face value of less than or equal to 50, and the non-change storage box is specifically used to store currencies with a face value of 100 that cannot be used for change;

[0123] The access control unit controls the classified storage of payment coins based on the face value mark and receives the change plan According to the change plan The change currency is extracted from the change inventory box in turn and delivered to the change channel in the currency transmission channel. After the change is completed, the recovery signal is fed back to the interactive display module. Based on the inventory change principle, a statistical decision is made whether to send a change inventory replenishment prompt or a full inventory prompt.

[0124] Furthermore, the access control unit statistically determines whether to send a change inventory replenishment prompt or a non-change full inventory prompt based on the inventory change principle, including:

[0125] When receiving payment real currency, deposit the payment real currency into the storage area corresponding to the currency face value in the cash box according to the face value mark, and change the currency inventory quantity sequence n x The amount of currency in stock corresponding to the currency denomination is increased by 1 to determine whether the amount of currency in stock corresponding to the currency denomination has reached the storage limit. If so, a full inventory notification is sent to the staff. The full inventory notification includes the currency denomination, the corresponding currency inventory quantity, and the amount of currency removed. If the storage limit is not reached, no action is taken.

[0126] When receiving the change proposal After the change is completed, re-count and generate the change currency inventory quantity sequence n x , execute the first proposition judgment to determine whether the total amount of change inventory currencies with denominations of 5 and 1 is greater than or equal to 9, and at the same time execute the second proposition judgment to determine whether the change inventory currencies with denominations of 50, 20 and 10 can be combined to generate all amounts less than 100 and multiples of 10. If both the first proposition judgment and the second proposition judgment are true, the change inventory currency can meet all change situations and no processing is performed. If the first proposition judgment and / or the second proposition judgment are not true, a change inventory replenishment reminder is sent to the staff, and the change inventory replenishment reminder includes the currency replenishment amount of each currency denomination.

[0127] Furthermore, the storage upper limit, currency removal quantity, and currency replenishment quantity involved in the inventory change principle are all related to the preset storage quantity of each currency denomination. There is a fixed ratio between the preset storage quantities of different currency denominations. The fixed ratio is pre-defined and negatively correlated with the currency denomination. The storage upper limits of different currency denominations are the same and greater than the preset storage quantity corresponding to any single currency denomination. The currency removal quantity corresponding to a single currency denomination is equal to the storage upper limit minus the preset storage quantity corresponding to the single currency denomination. The currency replenishment quantity corresponding to a single currency denomination is equal to the preset storage quantity corresponding to the single currency denomination minus the change currency inventory quantity corresponding to the single currency denomination. In this embodiment, the preset storage quantities corresponding to the currency denominations of 100, 50, 20, 10, 5, and 1 are 0, 4, 16, 40, 100, and 240 respectively, and the storage upper limit is 300.

[0128] Furthermore, the interactive display module includes a touch display screen, an induction encoding unit and a display control unit;

[0129] The touch screen uses capacitive touch technology to generate additional parasitic capacitance at the user's touch location and cause the local charge to increase. It receives the display image rendered by the display control unit and updates the display.

[0130] The induction encoding unit directly obtains the vehicle license plate through the scanning device, obtains the local charge peak based on the induction effective strategy, and determines whether to generate a confirmation signal in combination with the local charge increase area, and sends the vehicle license plate or confirmation signal to the central computing module;

[0131] When the display control unit receives display information, it adds or replaces it in the current rendering template. When it receives a restoration signal, it clears the display information in the current rendering template, renders and generates a display image, and sends it to the touch screen. Among them, the vehicle information and parking time are sequentially added to the table in a fixed position in the rendering template, and the total consumption amount, total payment amount and change amount respectively replace the values ​​in the corresponding positions in the rendering template.

[0132] like Figure 6 As shown, further, the sensing coding unit obtains the local charge peak value based on the sensing effective strategy and determines whether to generate a confirmation signal in combination with the local charge increase area, including the following specific steps:

[0133] Detecting a local charge increase, obtaining a local charge peak value, and determining whether the local charge peak value is greater than a touch effective valley value;

[0134] If the local charge peak value is less than or equal to the effective touch valley value, it is determined to be an invalid touch and the increase in local charge is ignored.

[0135] If the local charge peak value is greater than the touch effective valley value, the range of the local charge increase area on the touch screen is determined, compared with the range of the confirmation button, and the overlap ratio is calculated;

[0136] If the overlap ratio is greater than the overlap ratio threshold, it is determined to be a valid touch and a confirmation signal is generated. If the overlap ratio is less than or equal to the overlap ratio threshold, it is determined to be an invalid touch and the local charge increase is ignored.

[0137] The present invention discloses an AI multi-currency change parking control system, comprising a central computing module, a currency recognition module, a classification access module and an interactive display module; the central computing module accumulates the parking time based on the vehicle license plate and the confirmation signal and sends the total consumption amount, calculates the total payment amount and the total change amount through the cumulative change decision method, and generates a change plan; the currency recognition module obtains the payment currency grayscale image and adopts KL transformation and genetic evolution back propagation network to collaboratively identify the payment coin and the payment amount; the classification access module classifies and stores the payment coin and executes the change plan, and decides whether to send a change inventory replenishment prompt or a full inventory prompt based on the inventory change principle; the interactive display module senses the user operation to generate the vehicle license plate or the confirmation signal, and displays the display information in real time, thereby realizing an automated intelligent system integrating the three functions of accurate recognition, optimal change and automatic replenishment prompt.

[0138] The foregoing description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the foregoing embodiments. All technical solutions based on the principles of the present invention fall within the scope of protection of the present invention. It should be noted that improvements and modifications that are within the scope of protection of the present invention are within the scope of protection of the present invention.

Claims

1. An AI multi-currency change parking control system, characterized in that: It includes a central computing module, a currency identification module, a classification access module and an interactive display module; The central calculation module obtains vehicle information based on the vehicle license plate and accumulates the parking time, receives a confirmation signal and sends the total consumption amount, calculates the total payment amount and the total change amount through the cumulative change decision method, and generates a change plan; The currency recognition module obtains the payment currency grayscale image and uses KL transformation and genetic evolution back propagation network to collaboratively identify the payment coin and the payment amount; The classification access module stores payment real coins in categories, executes the change plan, and statistically decides whether to send a change stock replenishment reminder or a full stock reminder based on the inventory change principle; The interactive display module senses user operations to generate a vehicle license plate or confirmation signal, and displays information in real time. The display information includes vehicle information, parking time, total consumption amount, total payment amount and total change amount.

2. The AI ​​multi-currency change parking control system according to claim 1, characterized in that: The central computing module includes a main memory unit, a statistical unit and a change computing unit; The main memory unit stores a parking time table, and the parking time table can be queried to obtain vehicle information and parking time corresponding to the vehicle license plate; The statistical unit obtains and sends vehicle information and parking time based on the vehicle license plate to the interactive display module, accumulates the parking time to update the total consumption amount, and sends the total consumption amount to the interactive display unit when receiving the confirmation signal; The change calculation unit accumulates the payment amount based on the cumulative change decision method to obtain the total payment amount and the total change amount and sends them to the interactive display module, generates a change plan and sends it to the classification access module, and clears the total consumption amount and the total payment amount.

3. An AI multi-currency change parking control system as claimed in claim 1 or 2, characterized in that: The cumulative change decision-making method includes the following specific steps: Get the total amount of consumption and the inventory quantity of change currency, and set the total payment amount to 0; Receive the payment amount and add it to the current total payment amount to update the total payment amount, and determine whether the total payment amount is greater than or equal to the total consumption amount; If it is less than the total amount of consumption, wait for receiving the new payment amount; if it is greater than or equal to the total amount of consumption, calculate the total amount of change and send it together with the total amount of payment to the interactive display module; The change optimization goal is to minimize the total number of changes used. Set change optimization constraint 1 to ensure that the total amount of change currency used is equal to the total amount of change. Set change optimization constraint 2 to ensure that the number of change currencies of a single denomination used must be less than or equal to the corresponding change currency inventory. The change optimization objective, change optimization constraint 1 and change optimization constraint 2 are combined to generate a change optimization problem. The improved greedy algorithm is used to solve the problem and obtain the change solution, which is then sent to the classification storage module.

4. The AI ​​multi-currency change parking control system according to claim 3, characterized in that: The improved greedy algorithm is used to solve the change solution, which includes the following specific steps: Initialize the remaining total amount and change solution to the total change amount and all-zero vector; Execute change currency selection, use the change currency in stock corresponding to the largest currency denomination that is less than or equal to the current remaining total amount as the change currency, and determine whether the product of the change currency denomination and the change currency stock quantity is greater than the current remaining total amount; If it is greater than the current remaining total amount, the maximum number of change currencies that satisfies the requirement that the face value of the change currency multiplied by the number of change currencies used is less than or equal to the current remaining total amount will be used as the optimal number of change currencies used; If it is less than or equal to the current remaining total amount, the amount of change currency in stock will be used as the optimal amount of change currency to use; Place the optimal change currency usage number into the corresponding position of the change plan and update the remaining total amount to the current remaining total amount minus the product of the optimal change currency usage number and the change currency face value, and determine whether the remaining total amount is 0; If the remaining total amount is not equal to 0, jump to execute change currency selection. If the remaining total amount is equal to 0, obtain the change solution.

5. The AI ​​multi-currency change parking control system according to claim 1, characterized in that: The currency recognition module includes a scanning device, a feature extraction unit and an identification marking unit; The scanning module is used to obtain a grayscale image of the payment currency; The feature extraction module uses KL transformation to calculate the covariance matrix of the payment currency grayscale image and decomposes it to obtain eigenvalues, determines the contribution of the corresponding eigenvector according to the size of the eigenvalue, and generates effective features based on the contribution combination; The identification marking unit establishes a nonlinear mapping from effective features to currency authenticity and currency amount through a genetic evolution back propagation network. If it is a payment of real currency, the payment amount is sent to the central computing module and the face value mark is sent to the classification access module.

6. The AI ​​multi-currency change parking control system according to claim 5, characterized in that: The training of the genetic evolution back propagation network includes the following specific steps: Randomly generate the initial population, and use floating-point encoding to process each individual in the initial population to generate the corresponding chromosome; The inverse of the sum of squares of the error between the output value of the back propagation network and the actual value is used as the fitness function; The fitness proportional method and partial arithmetic crossover are used as the reproduction operator and crossover operator respectively, and the relationship between the mutation probability and the evolutionary generations is set to normal distribution to determine the mutation operator; Set the maximum evolutionary generation, calculate the fitness function values ​​of all individuals in the initial population, and determine whether the maximum evolutionary generation is met; If it is not satisfied, the two individuals with the largest reproduction operator are selected, and new individuals are generated based on the crossover operator and mutation operator. If the maximum evolutionary generation is met, the chromosome of the individual with the largest fitness function value in this generation of genetic evolution is retained as the threshold and connection weight of the back propagation network; Perform forward propagation of the backpropagation network, calculate the loss function value, and determine whether the loss function value is less than or equal to the threshold; If it is greater than the threshold, the gradient descent method is used to adjust the connection weights, recalculate the loss function value and continue to judge whether it is less than or equal to the threshold. If it is less than or equal to the threshold, the training is completed.

7. The AI ​​multi-currency change parking control system according to claim 1, characterized in that: The classification access module includes a cash box and an access control unit; The cash box is used to store currencies of different denominations in a classified manner; The access control unit stores payment real coins based on denomination marks, extracts change currency in sequence based on the change plan, feeds back a restoration signal to the interactive display module after the change is completed, and statistically decides whether to send a change stock replenishment prompt or a full stock prompt based on the inventory change principle.

8. The AI ​​multi-currency change parking control system according to claim 7, characterized in that: The access control unit statistically determines whether to send a change inventory replenishment prompt or a non-change full inventory prompt based on the inventory change principle, including: When the real currency is deposited into the cash box, the amount of currency in stock corresponding to the face value of the real currency in the change currency inventory quantity sequence is increased by 1, and the amount of currency in stock corresponding to the face value is judged to see whether the storage limit has been reached. If the storage limit is reached, a full storage prompt is sent; if the storage limit is not reached, no processing is performed; When the change is completed, the change currency inventory quantity sequence is re-counted and the first proposition judgment and the second proposition judgment are executed simultaneously to determine whether the change inventory currency can meet all the amounts from 1 to 99. If so, no processing is done. If not, a change inventory replenishment reminder is sent.

9. The AI ​​multi-currency change parking control system according to claim 1, characterized in that: The interactive display module includes a touch screen, an induction encoding unit and a display control unit; The touch display screen receives a display image and updates the display based on the increase of local charge caused by the user's touch; The sensing encoding unit obtains the vehicle license plate, decides whether to generate a confirmation signal based on the sensing validity strategy, and sends the vehicle license plate or the confirmation signal to the central computing module; The display control unit receives the display information and adds or replaces it to the current rendering template, clears the display information in the current rendering template upon receiving the restoration signal, renders and generates a display image and sends it to the touch display screen.

10. The AI ​​multi-currency change parking control system according to claim 9, characterized in that: The sensing encoding unit decides whether to generate a confirmation signal based on the sensing validity strategy, including the following specific steps: Detecting the increase of local charge, obtaining the local charge peak value and determining whether it is greater than the effective touch valley value; If it is less than or equal to the effective touch valley value, ignore the increase in local charge; If it is greater than the effective touch valley value, the range of the local charge increase area is confirmed, compared with the range of the confirmation button and the overlap ratio is calculated; If the coincidence ratio is greater than the coincidence ratio threshold, a confirmation signal is generated; if the coincidence ratio is less than or equal to the coincidence ratio threshold, the local charge increase is ignored.

Images