A dynamic parking recommendation method considering individual demand preferences of travelers

Abstract

This invention relates to a dynamic parking recommendation method that considers the individual needs and preferences of travelers, belonging to the field of smart parking technology. The method includes the following steps: first, acquiring information on travelers' parking needs and preferences and screening personalized alternative parking lots; then, establishing a dynamic parking recommendation model by comprehensively considering the interests of travelers and managers; and finally, evaluating the implementation effect of the smart parking recommendation system. This method can be used to provide travelers with personalized parking recommendations and reservation services during their trips, while also balancing the utilization of parking facility resources and maximizing overall benefits.

Description

Technical Field

[0001] This invention relates to a dynamic parking recommendation method that takes into account the individual needs and preferences of travelers. It belongs to the field of smart parking technology and can be used to provide travelers with personalized parking recommendation schemes and reservation services when parking, while also balancing the utilization of parking facility resources and maximizing overall benefits. Background Technology

[0002] With rising living standards, the number of cars owned has grown rapidly, exacerbating the parking problem. This has led to numerous instances of illegal parking and prolonged searching for parking spaces, significantly impacting road traffic and environmental pollution. Parking facilities are distributed across time and space; uneven utilization of these resources results in substantial waste. Intelligent parking services, which integrate information technology, communication technology, mobile terminal technology, and GIS technology to collect, manage, query, reserve, and guide parking spaces, reduce travelers' parking search time and maximize parking resource utilization efficiency and parking lot revenue. This represents an effective method for solving urban parking problems.

[0003] Currently, domestic and international methods for recommending parking mainly consider the attributes of parking lots, such as parking lot location, number of parking spaces, parking price, and walking distance from the parking lot to the destination. Based on the traveler's walking distance after parking, parking price, or number of parking spaces, the distribution of surrounding parking facilities is displayed for the traveler to choose and book. Examples include Jie Parking, ETCP Parking Management System, and Xiaoqiang Parking.

[0004] However, current parking recommendation methods mostly rely on one-way recommendations based on the lowest parking price or the shortest walking distance after parking. While this may alleviate some parking shortages, it still leads to significant resource imbalances between parking lots. Furthermore, many parking recommendation schemes assume that people have the same parking needs and preferences, but travelers often have different preferences and varying psychological thresholds for accepting parking-related factors. This leads to information redundancy or insufficiency in the recommendation process, creating information asymmetry, negatively impacting the traveler's parking experience, and ultimately affecting the effectiveness of parking recommendations and failing to achieve the goal of effectively balancing parking facility resource utilization. Moreover, current parking recommendation schemes often use fixed recommendation patterns without adjusting for different parking lot occupancy rates. They fail to prioritize maximizing traveler benefits when parking resources are abundant, and fail to balance resources among parking lots when they are scarce.

[0005] Therefore, it is necessary to consider travelers' personalized parking needs, preferences, and psychological characteristics, and to combine this with parking lot utilization to ensure a balanced parking experience while ensuring sufficient parking resources. This paper proposes a dynamic parking recommendation method that considers travelers' individual needs and preferences. When parking resources near the destination are sufficient, recommendations are made to maximize the traveler's benefit. When parking resources near the destination are uneven and scarce, parking resources are balanced to achieve the optimal combination of individual traveler benefits and parking management benefits, providing a reference for solving urban parking problems. Summary of the Invention

[0006] Based on the above analysis, this invention proposes a dynamic parking recommendation method that considers the individual needs and preferences of travelers. While ensuring the traveler's parking experience, it also considers the balanced use of parking resources and recommends the best parking lot to travelers in real time, thereby maximizing parking efficiency.

[0007] This invention starts from the individual parking needs and preferences of travelers. In the parking recommendation process, it obtains information such as the degree of attention and psychological threshold of individual travelers to parking-influencing factors, proposes a screening method for alternative parking lots that meet individual parking needs and preferences, then establishes a parking recommendation model that takes into account the individual parking needs and preferences of travelers, thereby providing parking lot recommendation schemes, and finally uses evaluation indicators to analyze the effectiveness of the parking recommendation schemes.

[0008] The invention is characterized by establishing a smart parking recommendation method based on the individual parking needs and preferences of travelers, taking into account the benefits of both travelers and managers. The resulting personalized parking recommendation scheme can not only meet individual needs and improve the individual travel and parking experience, but also achieve the effect of balancing the utilization of parking resources.

[0009] The technical concept of this invention is characterized by:

[0010] 1. Acquisition of travelers' parking needs and preferences and personalized selection of alternative parking lots.

[0011] 2. Taking into account the interests of both travelers and administrators, establish a dynamic parking recommendation model.

[0012] 3. Evaluation of the implementation effect of smart parking recommendation

[0013] To achieve the above objectives, the present invention employs the following steps:

[0014] Step 1. Obtaining traveler parking demand and preference information and filtering personalized alternative parking lots.

[0015] Step 1.1. Obtaining Travelers' Travel and Parking Needs and Preferences

[0016] When designing a parking recommendation system, GPS positioning technology can be used to obtain the traveler's origin information, as well as the traveler's destination location information and the information on all parking lots within a certain range around the destination, including the capacity of each parking lot, the number of parking spaces available in real time, parking prices, and the walking distance from the parking lot to the destination.

[0017] Simultaneously, it's crucial to obtain information on travelers' parking needs and preferences, including their level of attention to parking-related factors and their psychological thresholds. The level of attention travelers pay to parking factors refers to their level of awareness of these factors when making parking choices. The design method involves having travelers assess the importance of each parking-related factor during a specific parking trip. These factors include walking distance after parking and parking price, with importance options as follows: 1 represents very unimportant; 2 represents unimportant; 3 represents moderately important; 4 represents relatively important; and 5 represents very important. The psychological threshold for parking factors refers to the value at which travelers will abandon a parking lot if its attributes exceed their psychologically acceptable values. The design method provides numerical options for each parking-related factor, including walking distance after parking, parking price, and remaining parking spaces, allowing travelers to select their acceptable maximum or minimum value as their psychological threshold for a specific parking location.

[0018] Travelers can use the parking recommendation system's terminal app to input their travel destination information, then input their level of concern and psychological threshold regarding parking-related factors, and then view real-time parking information within a certain range of their destination.

[0019] Step 1.2. Personalized Alternative Parking Lot Filtering

[0020] Based on real-time parking information within a certain radius of the traveler's destination, and considering the traveler's level of attention and psychological threshold regarding parking factors (walking distance after parking, parking price, and remaining parking spaces), candidate parking lots that meet the traveler's psychological parking needs are selected. The selection method is as follows:

[0021] First, parking lots that simultaneously meet the psychological thresholds of the three factors affecting parking for travelers are selected as candidate parking lots. If the information of a certain parking lot j simultaneously meets the psychological thresholds of the three factors for travelers during this parking trip, it can be used as a candidate parking lot for traveler i, then Z ij =1, otherwise Z ij =0, expressed by formula (1).

[0022]

[0023] In the formula: C j Parking fee for parking lot j; L jThe walking distance from parking lot j to the destination; O tj Let TC be the number of empty spaces in parking lot j at time t; i The maximum acceptable parking fee for traveler i, i.e., the psychological threshold for parking price; TL i The maximum acceptable walking distance after parking for traveler i, i.e., the psychological threshold for walking distance after parking; TO i Let be the minimum number of parking spaces that traveler i can accept, i.e., the psychological threshold for the number of parking spaces.

[0024] If no parking lot information exists that simultaneously meets the psychological thresholds of the three factors for this parking trip, then parking lots that meet the psychological threshold of any one of the three factors will be selected as alternative parking lots that meet the traveler's parking psychological needs.

[0025]

[0026] If no parking lot meets the psychological threshold of any of the three factors, then all parking lots with available spaces within a certain range near the destination can be considered as alternative parking lots.

[0027] Following the above process, Z ij Parking lots with a value of 1 are selected as set A. ik This refers to alternative parking lots that meet the psychological parking needs of travelers during that particular trip.

[0028] Step 2. Taking into account the interests of travelers and administrators, establish a dynamic parking recommendation model.

[0029] It includes three parts: classification and standardization of parking influencing factors, construction of a weight matrix of factors influencing traveler parking, and establishment of a dynamic parking recommendation model that comprehensively considers the benefits for both travelers and managers.

[0030] Step 2.1. Classification and Standardization of Parking Influencing Factors

[0031] Based on the alternative parking lots obtained in step 1.2 that meet the psychological threshold of travelers' parking needs, parking influencing factors are classified and standardized from both the perspectives of travelers and managers.

[0032] When choosing a parking spot, travelers primarily consider the walking distance from the parking lot to their destination and the parking price. The closer a parking lot with available spaces is to the destination and the lower the parking price, the greater the benefit to the traveler. Therefore, these two factors are considered as traveler benefit factors. The traveler benefit factors are standardized as shown in formula (3), and the matrix is...

[0033]

[0034] In the formula: L ik C ik These represent the parking price and walking distance after parking for traveler i in the alternative parking lot k, respectively; ik c ik These are the standardized values ​​of the parking price and walking distance after parking for traveler i at alternative parking lot k, respectively; m i The number of alternative parking lots for traveler i.

[0035] For managers, in areas with high parking demand and insufficient parking spaces, the main consideration is the balanced and effective use of parking resources, focusing on the availability of parking lots. The goal is to guide travelers to parking lots with more vacant spaces to reduce long-term parking searches and haphazard parking. Therefore, parking vacancy is considered a managerial benefit factor, and a standardization method for managerial benefit factors is implemented, as shown in formula (4). Let represent the standardized value of the number of vacant spaces in the alternative parking lot k when recommending parking to traveler i from the manager's perspective. The matrix is...

[0036]

[0037] In the formula: O tik Let o be the number of available parking spaces for traveler i at time t in the alternative parking lot k. tik Let be the standardized value of the number of parking spaces available for traveler i at time t in the alternative parking lot k;

[0038] Step 2.2. Construct a weight matrix of factors influencing traveler parking.

[0039] By evaluating each traveler's level of attention to parking-related factors, the weight of each factor is calculated. This reflects the importance of individual traveler factors when making parking choices and also reveals the differences in individual traveler needs and preferences. Let traveler i's evaluation results for the three factors—walking distance from the parking lot to the destination, parking price, and number of available parking spaces—be as follows: Then the weight vector W of traveler i's attention to factors affecting parking i This is represented by formula (5).

[0040]

[0041] in,

[0042]

[0043]

[0044] in, The weights represent the degree of attention traveler i pays to the walking distance from the parking lot to the destination, the parking price, and the number of available parking spaces during this parking trip.

[0045] Step 2.3. Establish a dynamic parking recommendation model that comprehensively considers the benefits for both travelers and managers.

[0046] Considering both travelers and managers, the utility of candidate parking lots that meet the psychological threshold of travelers' parking needs is calculated, as shown in formula (6).

[0047]

[0048] In the formula: δ is an adjustment coefficient that considers the benefits or utility of travelers and managers. When δ = 1, it means that from the perspective of maximizing the benefits of travelers, the parking lot closer to the destination and with lower parking prices has greater utility. When δ = 0, it means that from the perspective of maximizing the benefits of managers, the parking lot with more available parking spaces has greater utility. When δ = (0,1), it considers both the benefits of travelers and the benefits of managers.

[0049] The utility value of each candidate parking lot for each traveler can be calculated according to formula (6). Based on the principle of maximizing utility, the recommended parking lot is obtained, as shown in formula (7).

[0050]

[0051] In the formula: Q ti Let t be the maximum utility of the alternative parking lots for traveler i at time t. This parking lot is the one recommended to traveler i.

[0052] To achieve balanced utilization of parking resources, a dynamic adjustment coefficient is adopted. Based on the utilization status of different parking lots and an adjustment threshold θ, parking recommendations are initiated and adjusted according to the needs of travelers, thereby achieving the goal of balanced utilization of parking resources.

[0053] The value of the dynamic adjustment coefficient δ is shown in formula (8).

[0054]

[0055] In the formula: Let be the median value of the occupancy rate of all parking lots near the destination of traveler i at time t, representing the average situation of the overall parking lot occupancy rate. Let be the average of all parking lots near traveler i's destination at time t, meaning that parking recommendations aim to balance and maintain a relatively uniform overall parking lot utilization rate. θ is the threshold for initiating the adjustment process.

[0056] When the median value of the occupancy rate of alternative parking lots At time t, it indicates that multiple parking lots near the destination are under high utilization and resources are relatively scarce. Therefore, a parking recommendation system is initiated, primarily focused on balancing parking resources, to guide more travelers to parking lots with more available spaces. At time t, it means that there are sufficient parking resources near the destination. At this time, the parking recommendation is mainly based on maximizing the interests of the traveler. The traveler can choose a parking lot that is closer to the destination and has a lower price.

[0057] The activation adjustment threshold θ can be set according to actual needs. If the threshold θ is set high, it means that parking recommendations based on maximizing manager benefits will only be activated when the overall utilization rate of each parking lot reaches a high value, thus ensuring balanced utilization of each parking lot. The activation adjustment time is generally longer. If the threshold θ is set low, it means that parking recommendations based on maximizing manager benefits will be activated when the overall utilization rate of each parking lot reaches a relatively low value, thus ensuring balanced utilization of each parking lot. The activation adjustment time is generally shorter.

[0058] Step 3. Evaluation of the implementation effect of smart parking recommendation

[0059] To analyze the effectiveness of the smart parking recommendation model in real time, multiple indicators need to be established for comprehensive evaluation. These indicators are mainly divided into two categories: traveler benefit evaluation indicators and manager benefit evaluation indicators. Traveler benefit evaluation indicators include average walking distance after parking, average parking fee, and the percentage of parking demand psychological thresholds met.

[0060] (1) Average walking distance after parking: The ratio of the total walking distance after parking for travelers who make parking selections based on parking recommendations to the total number of travelers.

[0061]

[0062] In the formula: L represents the average walking distance after parking. i Let n be the walking distance from the parking lot chosen by traveler i to their destination, and n be the total number of travelers who made the parking choice.

[0063] (2) Average parking cost: The ratio of the total parking cost of travelers who make parking choices based on parking recommendations to the total number of travelers.

[0064]

[0065] In the formula: For the average parking cost, f i The parking fee for the parking lot selected by traveler i.

[0066] (3) Parking demand psychological threshold satisfaction ratio: Based on parking recommendations, the ratio of all parking lots selected by travelers that meet the traveler's psychological thresholds for all factors, including walking distance after parking, parking price, and number of available parking spaces.

[0067] The proportion of travelers who meet all psychological thresholds for parking: the ratio of travelers who simultaneously meet the psychological thresholds for the three parking factors (parking price, walking distance after parking, and number of available parking spaces) to the total number of travelers.

[0068]

[0069] In the formula: B all The selected parking lot represents the number of people who simultaneously meet all psychological thresholds, where n is the total number of travelers.

[0070] Percentage of travelers whose primary concerns meet the psychological threshold for parking is the ratio of travelers whose selected parking lot meets only the psychological threshold for their primary concerns to the total number of travelers.

[0071]

[0072] In the formula: The number of travelers to meet the psychological threshold of their primary concerns regarding parking influencing factors.

[0073] Secondary concern factor psychological threshold satisfaction ratio: The ratio of travelers who meet only the secondary concern parking influencing factor psychological thresholds for the selected parking lot to the total number of travelers.

[0074]

[0075] In the formula: The number of travelers is required to meet the psychological threshold of the secondary parking-related factors.

[0076] The main performance indicators for managers include real-time parking occupancy rate and cumulative parking revenue.

[0077] (1) Real-time parking occupancy rate: the ratio of the number of parking spaces occupied in real time to the total capacity of parking spaces.

[0078]

[0079] In the formula: V k Let k be the total parking capacity of parking lot k. Let P be the number of empty spaces in parking lot k at time t. tk Let be the parking occupancy rate of parking lot k at time t.

[0080] (2) Cumulative parking revenue: This is the total parking fee revenue calculated based on the parking time after all travelers have selected a parking space according to the parking recommendation.

[0081]

[0082] In the formula: I represents the cumulative parking revenue.

[0083] To analyze the effectiveness of the smart parking recommendation model in real-time, based on the above evaluation indicators, multiple indicators need to be established for comprehensive evaluation. These indicators characterize and assess the dynamic parking recommendation scheme, which considers travelers' individual parking needs and preferences, in meeting travelers' parking demands and maintaining a balance in parking lot occupancy.

[0084] The inventiveness of this invention is mainly reflected in:

[0085] (1) Based on the traveler’s parking demand preference information, namely the degree of attention to parking influencing factors and psychological threshold, this invention proposes a personalized alternative parking lot screening method for travelers.

[0086] (2) This invention takes into account the benefits of travelers and managers. Based on the utilization of parking facilities, it proposes a dynamic parking recommendation method and evaluation method using dynamic adjustment coefficients and activation adjustment thresholds. This method can not only meet the needs of travelers and improve the parking experience, but also achieve the effect of balancing the utilization of parking resources. Attached Figure Description

[0087] Figure 1 The present invention provides a step-by-step diagram of a dynamic parking recommendation method that takes into account the individual needs and preferences of travelers.

[0088] Figure 2 Flowchart of the dynamic parking recommendation method of the present invention that takes into account the individual needs and preferences of travelers;

[0089] Figure 3 Optimal recommendation for travelers;

[0090] Figure 4 The manager's optimal recommendation solution;

[0091] Figure 5 Dynamic recommendation scheme. Detailed Implementation

[0092] The following is a specific implementation example of the present invention: Suppose that a traveler ID1 drives from home (Century Oriental City) to Xidan Joy City for shopping and dining, which takes about 3 hours and covers a distance of about 15 kilometers. There is a parking recommendation system terminal App that can query parking information around the destination. Based on this parking trip, the parking recommendation system recommends parking to traveler ID1.

[0093] Step 1. Obtaining traveler parking demand and preference information and filtering personalized alternative parking lots.

[0094] Step 1.1 Obtaining Travelers' Travel and Parking Needs and Preferences

[0095] Assuming information on parking availability within a certain radius of the destination, there are 5 parking lots within a 1km radius of Xidan Joy City, as shown in Table 1, with each parking lot having a total of 200 parking spaces.

[0096] Table 1: Parking information near the destination at time t

[0097] Parking lot number Parking price (RMB / hour) Walking distance after parking (m) Number of empty slots remaining P1 10 60 20 P2 10 100 30 P3 8 250 50 P4 8 520 100 P5 5 600 150

[0098] Information on travelers' parking needs and preferences, including their level of concern about factors affecting parking and their psychological thresholds.

[0099] (1) The degree of attention travelers pay to parking factors involves having travelers assess the importance of various parking-related factors (walking distance after parking, parking price, number of available parking spaces) based on a specific parking trip. The design includes:

[0100] Based on this parking trip, how much do you care about the walking distance from the parking lot to your destination? A. Very unimportant B. Unimportant C. Generally important D. Quite important E. Very important

[0101] Based on your level of concern regarding parking fees for this parking trip: A. Very unimportant B. Unimportant C. Moderately important D. Quite important E. Very important

[0102] Based on your parking situation, how much do you care about the remaining parking spaces? A. Very unimportant B. Unimportant C. Moderately important D. Quite important E. Very important

[0103] The importance level is assigned from very unimportant to very important, with values ​​ranging from 1 to 5.

[0104] (2) Traveler's psychological threshold for parking factors: This involves providing numerical values ​​for each parking factor and allowing travelers to select their psychological threshold for a given parking trip. The designed questions include:

[0105] Based on this parking trip, at what walking distance after parking will you no longer consider choosing this parking lot: 1000 meters, 800 meters, 500 meters, 300 meters, 200 meters, 100 meters.

[0106] Based on the parking fee for this trip, at what price will you no longer consider choosing this parking lot: 20 yuan / h, 15 yuan / h, 10 yuan / h, 8 yuan / h.

[0107] Based on the number of available parking spaces for this parking trip, you will no longer consider selecting this parking lot: 30, 20, 10, 5, 2.

[0108] Taking traveler ID1 as an example, the survey results show that their level of concern for walking distance to parking is 5 (very important), and their level of concern for parking price is 3 (generally important). Their acceptable threshold for walking distance after parking is 1000m, their psychological threshold for parking price is 8 yuan / hour, and their psychological threshold for the number of available parking spaces is 2.

[0109] Step 1.2. Personalized Alternative Parking Lot Filtering

[0110] Based on the parking lots within 1km of the traveler's destination, Xidan Joy City, and combined with the traveler ID1's attention to parking factors and psychological threshold, P3, P4, and P5 were selected as candidate parking lots that meet the traveler ID1's parking psychological needs at time t. They simultaneously satisfy the traveler's psychological threshold for the three factors affecting parking, as shown in Table 2.

[0111] Table 2 shows the personalized alternative parking lots for traveler ID1 at time t.

[0112]

[0113] Step 2. Taking into account the interests of travelers and administrators, establish a dynamic parking recommendation model.

[0114] Step 2.1. Classification and Standardization of Parking Influencing Factors

[0115] Based on the alternative parking lots obtained in step 1.2 that meet the psychological threshold of traveler ID1's parking needs, parking influencing factors are classified and standardized from both the perspectives of travelers and managers. For traveler benefit factors, walking distance and parking price after parking are considered, and for manager benefit factors, the number of parking spaces is considered. Standardization is performed according to formula (3), and the results of standardization are shown in Table 3.

[0116] Table 3 Standardized Attribute Factors of Alternative Parking Lots for Traveler ID1 at Time t

[0117]

[0118] Therefore, the standardized matrix of the candidate parking lot attribute factors for traveler ID1 is obtained, and the standardized matrix R for traveler benefit factors is obtained. 13 Standardized matrix G of traveler benefit factors t13 .

[0119]

[0120] Step 2.2 Weight Matrix of Factors Affecting Traveler Parking

[0121] Based on the evaluation results of traveler ID1's attention to parking-related factors, their attention to walking distance after parking was assigned as "very important" with a value of 5, and their attention to parking price was assigned as "generally important" with a value of 3. The weight vector W of traveler ID1's attention to parking-related factors was calculated according to formula (5). i .

[0122]

[0123] Step 2.3 Dynamic parking recommendation model that comprehensively considers the benefits for both travelers and managers

[0124] Considering both travelers and managers, the utility of alternative parking lots that meet the psychological threshold of traveler ID1's parking demand is calculated according to formula (6). The adjustment coefficient δ is dynamically adjusted. Since a parking resource utilization rate of 60-80% generally indicates moderate parking resource utilization, the adjustment threshold θ is set to 70%. When <70%, recommendations are made to maximize traveler benefits; when ≥70%, parking resources are balanced. The median parking occupancy rate within 1km of the destination at time t is calculated according to formulas (6)(7)(8). A parking occupancy rate of 75%, exceeding the adjustment threshold of 70%, indicates a relatively scarce parking resource within the destination area. A dynamic recommendation scheme can be used to balance parking resources and recommend parking options. At this point, the average parking occupancy rate of all parking lots is... The value is 65%, i.e., δ = 0.35. Therefore, the utility of the alternative parking lots is:

[0125]

[0126] Parking recommendations are made based on maximizing utility. Parking lots are ranked according to their utility as P5, P4, and P3. Parking lot P5, which has the highest utility, is recommended as the parking lot for traveler ID1 at time t.

[0127] To compare and analyze the impact of different adjustment coefficient values ​​on parking recommendation results, we take two cases, δ=1 and δ=0, respectively, which represent parking recommendations based entirely on maximizing the benefits for travelers and maximizing the benefits for managers, and compare the results with those under the dynamic adjustment coefficient.

[0128] When δ = 1, the utility of the alternative parking lots for traveler ID1 is:

[0129]

[0130] When δ = 0, the utility of the alternative parking lots for traveler ID1 is:

[0131]

[0132] Therefore, when δ = 1, the order of utility from highest to lowest, based on maximizing utility, is P3, P5, P4. Thus, P3 is recommended to travelers first when δ = 1. When δ = 0, the order of utility from highest to lowest is P5, P4, P3. Therefore, P5 is recommended to travelers first when δ = 0. Thus, based on maximizing traveler benefit, P3 is recommended. However, if considering the equitable distribution of parking resources, P5 is recommended.

[0133] Step 3. Evaluation of the implementation effect of smart parking recommendation

[0134] Based on the dynamic parking recommendation process for a single traveler, the application is extended to multiple travelers. It is assumed that car travelers are randomly generated within a 5-30km range from the Xidan Joy City shopping mall, with a generation rate following a Poisson distribution with a mean of 23 vehicles / 5 minutes, and all destinations are Xidan Joy City. The parking time distribution of car travelers in the shopping mall, based on field survey data, is as follows: ≤0.5h 4%, 0.5-1h 8%, 1-2h 16%, 2-3h 32%, 3-4h 22%, and ≥4h 18%. Vehicles automatically leave the parking lot after reaching their parking time. Initially, vehicles in the parking lot leave at a frequency of 3 vehicles / 5 minutes. An adjustment threshold θ = 70% is set, and the adjustment coefficient δ is set to three different recommendation schemes: a dynamically changing adjustment coefficient based on parking lot utilization, δ = 1, and δ = 0. The implementation effects are compared and analyzed. The simulation duration is 270 minutes. Table 4 shows the parking lot information near the initial destination.

[0135] Table 4. Parking information near the initial destination

[0136] Parking lot number Parking price (RMB / hour) Walking distance after parking (m) Number of parking spaces P1 10 60 20 P2 10 100 80 P3 8 250 80 P4 8 520 100 P5 5 600 100

[0137] From the perspective of traveler benefit evaluation indicators, these include average walking distance after parking, average parking cost, and the proportion of parking demand psychological threshold satisfaction, as shown in Table 5.

[0138] Table 5. Evaluation Indicators for the Implementation Effectiveness of Different Parking Recommendation Schemes

[0139]

[0140] Table 5 shows that under the manager-optimal and dynamic parking recommendation schemes, the proportion of travelers meeting the psychological thresholds for all factors and the initial focus factors is greater than or equal to 90%, while it is relatively lower under the traveler-optimal parking recommendation scheme. Regarding the average walking distance after parking, the traveler-optimal parking recommendation scheme is the lowest at 540m. The average parking cost is roughly the same across all schemes.

[0141] From the perspective of management performance evaluation indicators, these include real-time parking occupancy rate and cumulative parking revenue. For example... Figure 3 As shown

[0142] from Figure 3 It can be seen that, from the perspective of maximizing traveler benefits, parking recommendations are δ=1. Generally, travelers first choose parking lots closer to their destination, and only choose parking further away when those nearby parking lots are full. Although the walking distance after parking at P5 is slightly longer than that at P4, the parking price at P5 is lower, resulting in higher overall utility. Therefore, P5 has a higher selection rate.

[0143] Figure 4 The data shows that parking recommendations are made based on maximizing the benefits for managers, i.e., δ=0. This mainly aims to balance the utilization of parking resources. Parking regulation is initiated from the beginning, and after a period of time, the occupancy rate of each parking lot can be kept at a basically the same level, changing synchronously with the increase or decrease of parking demand.

[0144] Figure 5 The data shows that in the initial stage of the dynamic parking recommendation scheme, since the median occupancy rate of parking lots near the destination is 60%, which is less than the adjustment threshold of 70%, parking recommendations are made based on the traveler's preference. Overall, the occupancy rate of parking lots closer to the destination gradually increases. When about 65 minutes have passed, the median occupancy rate of parking lots near the destination reaches 70%. At this point, parking recommendations are initiated based on balancing parking resources, guiding more travelers to park in parking lots with more available spaces, and the occupancy rates of various parking lots gradually move towards a balanced level.

[0145] As can be seen from Table 5, the cumulative parking revenue is not significantly different among the different parking recommendation schemes.

[0146] In summary, while recommending parking based on the traveler's optimal needs results in an imbalance between saturated nearby parking lots and vacant spaces in more distant ones, leading to a lower average walking distance. Recommending parking based on the administrator's optimal needs involves initiating system-wide parking adjustments from the outset, gradually achieving a balance in parking utilization across all parking lots. The dynamic parking recommendation scheme balances the interests of both travelers and administrators. When parking occupancy is low, it prioritizes maximizing traveler utility, meeting a higher proportion of travelers' needs and psychological thresholds. As parking demand increases and parking occupancy becomes high, it initiates system-wide balancing adjustments, ensuring balanced resource utilization across all parking lots and reducing prolonged parking searches. This is the optimal parking recommendation scheme.

[0147] The above are typical embodiments of the present invention, and the implementation of the present invention is not limited thereto.

Claims

1. A dynamic parking recommendation method that considers travelers' individual needs and preferences, characterized in that: Step 1. Obtaining traveler parking demand and preference information and filtering personalized alternative parking lots. Step 1.

1. Obtaining Travelers' Travel and Parking Needs and Preferences Using GPS positioning technology, information such as the traveler's departure point, destination, and parking availability within a certain radius of the destination can be obtained. At the same time, it is necessary to obtain information on travelers' parking needs and preferences, including factors influencing parking, level of attention, and psychological thresholds; Step 1.

2. Personalized Alternative Parking Lot Filtering First, parking lots that simultaneously meet the psychological thresholds of the three factors affecting parking for travelers are selected as candidate parking lots. If the information of a certain parking lot j simultaneously meets the psychological thresholds of the three factors for travelers during this parking trip, then it is selected as a candidate parking lot for traveler i, and Z... ij =1, otherwise Z ij =0, expressed by formula (1); （1） In the formula: C j Parking fee for parking lot j; L j The walking distance from parking lot j to the destination; O tj Let be the number of empty spaces in parking lot j at time t; TC i The maximum acceptable parking fee for traveler i, i.e., the psychological threshold for parking price; TL i The maximum acceptable walking distance after parking for traveler i, i.e., the psychological threshold for walking distance after parking; TO i This represents the minimum number of parking spaces that traveler i can accept, i.e., the psychological threshold for the number of parking spaces. Z ij Parking lots with a value of 1 were selected as set A. ik That is, as an alternative parking lot that meets the psychological parking needs of travelers during this trip; Step 2. Taking into account the interests of travelers and administrators, establish a dynamic parking recommendation model. Step 2.

1. Classification and Standardization of Parking Influencing Factors Based on the alternative parking lots obtained in steps 1 and 2; Standardize the influencing factors of traveler benefits, as shown in formula (3), the matrix is ​​as follows: ; （3） In the formula: , These represent the parking price and walking distance after parking for traveler i in the alternative parking lot k, respectively. These are the standardized values ​​of the parking price and walking distance after parking for traveler i in the alternative parking lot k, respectively. The number of alternative parking lots for traveler i; The standardization method for the parking space factor is shown in formula (4); This represents the standardized number of available parking spaces in candidate parking lots k when recommending parking spaces to traveler i from a manager's perspective. The matrix is ​​as follows: ; （4） In the formula: Let t be the number of available parking spaces for traveler i in the alternative parking lot k at time t. Let be the standardized value of the number of parking spaces available for traveler i at time t in the alternative parking lot k; Step 2.

2. Construct a weight matrix of factors influencing traveler parking. Let traveler i's evaluation results regarding their attention to the three factors—walking distance from the parking lot to their destination, parking price, and number of available parking spaces—be as follows: , , Then, the weight vector W for traveler i's attention to factors affecting parking is... i Represented as formula (5); （5） in, in, , , The weights of traveler i’s attention to the walking distance from the parking lot to the destination, the parking price, and the number of available parking spaces during this parking trip are used to indicate the importance of these factors. Step 2.

3. Establish a dynamic parking recommendation model that comprehensively considers the benefits for both travelers and managers. As shown in formula (6); In the formula: δ is an adjustment coefficient that considers the benefits or utility of travelers and managers; when δ=1, it means that from the perspective of maximizing the benefits of travelers, the parking lot that is closer to the destination and has a lower parking price has greater utility; when δ=0, it means that from the perspective of maximizing the benefits of managers, the parking lot with more empty parking spaces has greater utility; when δ=(0, 1), it considers both the benefits of travelers and the benefits of managers. The utility value of each candidate parking lot for each traveler is calculated according to formula (6). Based on the principle of maximizing utility, the recommended parking lot is obtained, as shown in formula (7). （7） In the formula: Let t be the maximum utility of the alternative parking lots for traveler i at time t, and that parking lot is the one recommended to traveler i. Based on the utilization of different parking lots, an adjustment threshold is applied. By recommending parking options, the system adjusts parking recommendations based on the needs of travelers, thereby achieving the goal of balanced utilization of parking resources across different parking lots. The dynamic adjustment proportional coefficient δ is set as shown in formula (8); （8） In the formula: Let be the median value of the occupancy rate of all parking lots near the destination of traveler i at time t; Let be the average of all parking lots near traveler i's destination at time t; To initiate the adjustment threshold; When the median value of the occupancy rate of alternative parking lots > When, it indicates that multiple parking lots near the destination have high utilization rates at time t; when ≤ At time t, it means that there are sufficient parking resources near the destination. At this time, parking recommendations are made to maximize the interests of travelers. Step 3. Evaluation of the implementation effect of smart parking recommendation To analyze the implementation effect of the above-mentioned intelligent parking recommendation model in real time, multiple indicators are established for comprehensive evaluation. These indicators are divided into two categories: traveler benefit evaluation indicators and manager benefit evaluation indicators. The traveler benefit evaluation indicators include average walking distance after parking, average parking fee, and the proportion of parking demand psychological threshold satisfied. The manager benefit evaluation indicators include real-time parking occupancy rate and cumulative parking revenue.

2. The method according to claim 1, characterized in that: Parking information includes the capacity of each parking lot, the number of available parking spaces in real time, parking prices, and the walking distance from the parking lot to the destination.

3. The method according to claim 1, characterized in that: Travelers can use the parking recommendation system's terminal app to input their travel destination information, then input their level of concern and psychological threshold regarding parking-related factors, and then view parking information within a certain range of their destination in real time.

4. The method according to claim 1, wherein step 3 specifically comprises: (1) Average walking distance after parking: The ratio of the total walking distance after parking for travelers who made parking selections based on parking recommendations to the total number of travelers; （9） In the formula: L represents the average walking distance after parking. i Let n be the walking distance from the parking lot chosen by traveler i to the destination, and n be the total number of travelers who made the parking choice. (2) Average parking cost: The ratio of the total parking cost of travelers who make parking choices based on parking recommendations to the total number of travelers; （10） In the formula: For the average parking cost, f i Parking fees for the parking lot selected by traveler i; (3) Parking demand psychological threshold satisfaction ratio: Based on parking recommendations, the ratio of parking lots selected by all travelers that meet the traveler's psychological thresholds for all factors, including walking distance after parking, parking price, and number of available parking spaces; The proportion of travelers who meet all psychological thresholds for parking: the ratio of travelers who simultaneously meet the psychological thresholds for the three parking factors (parking price, walking distance after parking, and number of available parking spaces) to the total number of travelers. （11） In the formula: The number of people who simultaneously meet all psychological thresholds for the selected parking lot, where n is the total number of travelers; Percentage of travelers whose primary concerns about parking are satisfied: The ratio of travelers whose primary concerns about parking are satisfied only for the selected parking lot to the total number of travelers. （12） In the formula: B first The number of travelers required to meet the psychological threshold of their primary concerns regarding parking factors; Secondary concern factor psychological threshold satisfaction ratio: The ratio of the number of travelers who meet only the secondary concern parking influencing factor psychological threshold for the selected parking lot to the total number of travelers; （13） In the formula: B second The number of travelers who meet the psychological threshold of the secondary parking-related factors. The performance evaluation indicators for managers include real-time parking occupancy rate and cumulative parking revenue; (1) Real-time parking occupancy rate: the ratio of the number of parking spaces occupied in real time to the total capacity of parking spaces; （14） In the formula: V k Let k be the total parking capacity of parking lot k. Let P be the number of empty spaces in parking lot k at time t. tk Let be the parking occupancy rate of parking lot k at time t; (2) Cumulative parking revenue: This is the total parking fee revenue calculated based on the parking time after all travelers have selected a parking space according to the parking recommendation. （15） In the formula: I represents the cumulative parking revenue; Based on the above evaluation indicators, in order to analyze the implementation effect of the above intelligent parking recommendation model in real time, multiple indicators are established for comprehensive evaluation; to characterize and evaluate the dynamic parking recommendation scheme that takes into account the individual parking needs and preferences of travelers in meeting travelers' parking needs and the balance of occupancy rates among parking lots.

5. The method according to claim 1, characterized in that: If there are no parking lots that simultaneously meet the psychological thresholds of the three factors for this parking trip, then parking lots that meet the psychological threshold of any one of the three factors will be selected as alternative parking lots that meet the traveler's parking psychological needs. （2） If no parking lot meets the psychological threshold of any of the three factors, then all parking lots with available spaces within a certain range near the destination can be considered as alternative parking lots.

Images