A parking lot operation optimization method and system based on multi-dimensional analysis

Abstract

This invention relates to a parking lot operation optimization method and system based on multi-dimensional analysis. The method filters similar parking lot clusters by calculating the similarity between the first multi-dimensional data of the parking lot to be optimized and the second multi-dimensional data of the parking lots to be selected. From these clusters, the parking lot with the highest operating index (derived from actual operating value and maximum operating potential value) is selected as a benchmark parking lot. The benchmark key indicators of the benchmark parking lot are compared with the key indicators to be optimized of the parking lot to be optimized. Based on the differences in key indicators and the actual operational resource constraints of the parking lot to be optimized, feasible optimization suggestions are generated to achieve optimization of the parking lot to be optimized. Therefore, this invention not only achieves precise and scientific selection of comparison targets for parking lots to be optimized, but also ensures that the generated optimization suggestions are not only compatible with the parking lot to be optimized but also feasible to implement.

Description

Technical Field

[0001] This invention relates to the field of parking lot operation technology, and in particular to a parking lot operation optimization method and system based on multi-dimensional analysis. Background Technology

[0002] Currently, parking lot operation optimization has become a core focus of the industry. When a parking lot is performing poorly, managers often rely on personal experience or roughly refer to data from individual parking lots in similar locations and of similar size to conduct a qualitative analysis of the parking lot to be optimized in order to generate optimization suggestions.

[0003] This method has several significant drawbacks: firstly, it relies excessively on human experience in selecting comparison targets, leading to strong subjectivity and low efficiency; secondly, using only geographical location and parking lot size as screening dimensions is too one-sided and lacks accuracy. These shortcomings directly result in the lack of scientific rigor and comparability of the selected comparison targets, causing optimization suggestions based on these targets to often mismatch with the actual situation of the parking lot to be optimized. This not only makes it difficult to achieve the expected optimization results but may even result in optimization suggestions that cannot be implemented. Summary of the Invention

[0004] The technical problem to be solved by this invention is: This invention provides a parking lot operation optimization method and system based on multi-dimensional analysis, which can accurately and scientifically screen the comparison objects of the parking lots to be optimized, while ensuring that the generated optimization suggestions match the parking lots to be optimized, and that the optimization suggestions can be implemented, so as to achieve effective optimization of the parking lots to be optimized.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] In a first aspect, the present invention provides a parking lot operation optimization method based on multi-dimensional analysis, comprising:

[0007] Based on the constructed multi-dimensional indicator system, the first multi-dimensional data of the parking lot to be optimized and the second multi-dimensional data of the parking lot to be selected are collected. The similarity between the first multi-dimensional data and the second multi-dimensional data is calculated. All parking lots to be selected with a similarity exceeding a first similarity threshold are regarded as similar parking lot clusters. The multi-dimensional indicator system includes basic indicator attributes, operational indicator attributes and environmental indicator attributes.

[0008] Obtain the actual operating value and maximum operating potential value of each parking lot to be selected in the similar parking lot cluster. Calculate the corresponding operating index based on the actual operating value and maximum operating potential value of each parking lot to be selected. Use the parking lot with the highest operating index as the benchmark parking lot for the parking lot to be optimized.

[0009] Based on the constructed key indicator system, benchmark key indicators of the benchmark parking lot and key indicators to be optimized of the parking lot to be optimized are collected. The key indicators to be optimized are compared with the benchmark key indicators to generate difference key indicators.

[0010] Based on the aforementioned key indicators of difference and the actual operational resource constraints of the parking lot to be optimized, feasible optimization suggestions are generated, and the parking lot to be optimized is optimized according to the optimization suggestions.

[0011] The beneficial effects of this invention are as follows: By constructing basic indicator attributes, operational indicator attributes, and environmental indicator attributes, a cluster of similar parking lots to the parking lot to be optimized is selected. This breaks through the overly one-sided nature of existing technologies in selecting comparison targets, while achieving precise and scientific selection. Benchmark parking lots are selected based on the operational index obtained from the actual operating value and maximum operating potential value of the parking lots to be selected, ensuring that the selected comparison targets have real and referable value, further improving the feasibility of the selection. By comparing the key indicators of the benchmark with the key indicators to be optimized, difference key indicators are generated to generate optimization suggestions. This allows the optimization of the parking lot to be optimized to focus on the core differences. At the same time, the actual operational resource constraints of the parking lot to be optimized are incorporated when generating optimization suggestions, avoiding the generated optimization suggestions from being divorced from reality, improving the matching degree and feasibility of the optimization suggestions with the parking lot to be optimized, and enabling the parking lot to be optimized to achieve effective optimization.

[0012] Optionally, the first multi-dimensional data includes first basic indicator data, first operational indicator data, and first environmental indicator data; the second multi-dimensional data includes second basic indicator data, second operational indicator data, and second environmental indicator data; and calculating the similarity between the first multi-dimensional data and the second multi-dimensional data, and classifying all candidate parking lots whose similarity exceeds a first similarity threshold as a cluster of similar parking lots, includes:

[0013] The basic similarity between the first basic indicator data and the second basic indicator data is calculated using a first similarity formula. Simultaneously, the operational similarity between the first operational indicator data and the second operational indicator data is calculated using a second similarity formula. Finally, the environmental similarity between the first environmental indicator data and the second environmental indicator data is calculated using a third similarity formula. The first similarity formula is as follows:

[0014] ;

[0015] in, This represents the basic similarity between parking lot i to be optimized and parking lot j to be selected. Indicates the similarity of parking lot types. Indicates the similarity of business types. Indicates the similarity in parking space size. This indicates the weights affected by the similarity of parking lot types. This indicates the weighting influenced by the similarity of business types. This indicates the weight influenced by the similarity in parking space size;

[0016] The second similarity formula is:

[0017] ;

[0018] in, This represents the operational similarity between parking lot i to be optimized and parking lot j to be selected. Indicates the similarity of rate structures. Indicates the similarity of operating duration. Indicates the similarity of monthly rent ratios. This indicates the weights affected by the similarity of the rate structure. This indicates the weighting influenced by the similarity of operating duration. This indicates the weight influenced by the similarity of monthly rent ratios;

[0019] The third similarity formula is:

[0020] ;

[0021] in, This represents the environmental similarity between parking lot i to be optimized and parking lot j to be selected. Indicating the similarity of subway environments, Indicates the similarity of bus routes. Indicates the environmental similarity of main roads. This indicates the weight influenced by the similarity of the subway environment. This indicates the weights affected by the similarity of bus routes. This indicates the weight influenced by the similarity of the main road environment;

[0022] The basic similarity, the operational similarity, and the environmental similarity are input into the total similarity formula for calculation to obtain the total similarity, wherein the total similarity formula is:

[0023] ;

[0024] in, This represents the total similarity between parking lot i to be optimized and parking lot j to be selected. This represents the weights that influence the basic similarity. This indicates the weight that influences the similarity of operations. This represents the weight that influences environmental similarity;

[0025] All parking lots to be selected whose total similarity exceeds the first similarity threshold are considered as a similar parking lot cluster.

[0026] As described above, the total similarity is obtained by weighted fusion of basic similarity, operational similarity, and environmental similarity, achieving hierarchical quantitative evaluation of similarity and avoiding misjudgments caused by single-dimensional calculation. Furthermore, specific calculation formulas have been designed for basic similarity, operational similarity, and environmental similarity, and each dimension of similarity has been further refined into sub-dimensions and the weights affected by the sub-dimensions, enabling dynamic calculation of the similarity of each dimension and improving the accuracy of the obtained similarity scores, thereby improving the accuracy of the obtained similar parking lot clusters.

[0027] Optionally, obtaining the actual operating value and maximum operating potential value of each candidate parking lot in the similar parking lot cluster, and calculating the corresponding operating index based on the actual operating value and maximum operating potential value of each candidate parking lot includes:

[0028] The parking space turnover rate and parking space utilization rate of each candidate parking lot in the similar parking lot cluster are calculated during the daytime and nighttime periods, respectively, to obtain the daytime turnover rate, nighttime turnover rate, daytime utilization rate and nighttime utilization rate of each candidate parking lot. The daytime revenue ratio of each candidate parking lot is calculated based on the daytime turnover rate and daytime utilization rate of each candidate parking lot. At the same time, the nighttime revenue ratio of each candidate parking lot is calculated based on the nighttime turnover rate and nighttime utilization rate of each candidate parking lot.

[0029] The actual daytime operating value of each parking lot is calculated based on its daytime turnover rate, daytime utilization rate, and daytime revenue ratio. At the same time, the actual nighttime operating value of each parking lot is calculated based on its nighttime turnover rate, nighttime utilization rate, and nighttime revenue ratio.

[0030] Obtain the daytime operating hours and average daytime parking time corresponding to the daytime utilization rate of each candidate parking lot. At the same time, obtain the nighttime operating hours and average nighttime parking time corresponding to the nighttime utilization rate of each candidate parking lot. Calculate the potential daytime turnover rate of each candidate parking lot based on its daytime utilization rate, daytime operating hours, and average daytime parking time. Calculate the potential nighttime turnover rate of each candidate parking lot based on its nighttime utilization rate, nighttime operating hours, and average nighttime parking time.

[0031] The maximum utilization rate during the daytime period is selected from the daytime utilization rate of each parking lot to be selected as the potential value of daytime utilization. At the same time, the maximum utilization rate during the nighttime period is selected from the nighttime utilization rate of each parking lot to be selected as the potential value of nighttime utilization.

[0032] The maximum daytime operating potential is calculated based on the daytime turnover rate potential and daytime utilization rate potential of each parking lot to be selected. At the same time, the maximum nighttime operating potential is calculated based on the nighttime turnover rate potential and nighttime utilization rate potential of each parking lot to be selected. The ratio of the actual daytime operating value of each parking lot to the maximum daytime operating potential is used as the daytime operating index of each parking lot to be selected. The ratio of the actual nighttime operating value of each parking lot to the maximum nighttime operating potential is used as the nighttime operating index of each parking lot to be selected.

[0033] The final operating index for each parking lot to be selected is obtained based on its daytime and nighttime operating indices.

[0034] As described above, the operating index is divided into daytime and nighttime operating indices to reflect the actual operating conditions of the parking lots to be selected. Different operating indices are calculated based on their corresponding actual operating values ​​and maximum operating potential values. The actual operating value incorporates turnover rate, utilization rate, and accounts receivable ratio, providing a comprehensive and accurate assessment of actual operating performance. The maximum operating potential value is generated based on the turnover rate potential value and utilization rate potential value, which better reflects the objective conditions of the parking lots to be selected and is therefore more reasonable. Ultimately, this improves the rationality and accuracy of the final operating index.

[0035] Optionally, the step of generating feasible optimization suggestions based on the key difference indicators and the actual operational resource constraints of the parking lot to be optimized, and optimizing the parking lot to be optimized according to the optimization suggestions, includes:

[0036] Based on the aforementioned key difference indicators, a linear regression model is constructed that relates to the operating index of the parking lot to be optimized. The linear regression model is as follows:

[0037]

[0038] in, This indicates the operational index of parking lots that need optimization. This represents the current value of the nth key difference indicator. This represents the weight of the nth key difference indicator on the operating index. Represents the residual. Indicates the total number of key indicators of difference;

[0039] Obtain the operating index of the benchmark parking lot, and simultaneously calculate the current index difference between the operating index of the benchmark parking lot and the operating index of the parking lot to be optimized. Based on the current index difference and the linear regression model, construct a first correlation constraint, which is:

[0040] ;

[0041] in, This represents the current value of the key indicator for the nth benchmark parking lot. This indicates the operating index of benchmark parking lots;

[0042] Using the minimum optimization cost of the parking lot to be optimized as the objective function, and under the first association constraint and the actual operating resource constraint of the parking lot to be optimized, a multi-round linear programming simulation is used to generate an executable optimization suggestion. The optimization of the parking lot to be optimized is realized according to the optimization suggestion, wherein the optimization suggestion includes the optimization target value of each difference key indicator.

[0043] The objective function is:

[0044] ;

[0045] ;

[0046] in, Let represent the objective function of the parking lot to be optimized. This represents the current value of the nth key difference indicator. This represents the current value of the key indicator for the nth benchmark parking lot. This represents the optimization cost of the nth key difference indicator. The cost of investing in optimizing the nth key difference indicator. The time cost of optimizing the nth key difference indicator;

[0047] The actual operational resource constraints include: indicator upper limit constraints, second correlation constraints, and resource upper limit constraints, wherein the indicator upper limit constraints are:

[0048] ;

[0049] in, This represents the upper limit of improvement for the nth key difference indicator;

[0050] The second association constraint is:

[0051] ;

[0052] in, This represents the lower limit threshold of the association between the nth key difference indicator and the mth key difference indicator. This represents the correlation coefficient between the nth difference key indicator and the nth benchmark key indicator. This represents the correlation coefficient between the m-th difference key indicator and the m-th benchmark key indicator. This represents the upper limit threshold for the association between the nth differential key indicator and the mth differential key indicator;

[0053] The resource upper limit constraint is:

[0054] ;

[0055] in, This indicates the maximum available investment in the parking lot to be optimized. This indicates the maximum available time for the parking lot to be optimized.

[0056] As described above, by constructing a linear regression model, the quantitative relationship between the key difference indicators and the operating indicators of the parking lot to be optimized is clarified, making the optimization suggestions more targeted. Based on the linear regression model and the current index difference between the operating index of the benchmark parking lot and the operating index of the optimized parking lot, the first correlation constraint is constructed to ensure that in subsequent multi-round linear programming simulations of the objective function, the focus remains on catching up with the benchmark parking lot and improving the operating indicators of the parking lot to be optimized, thus avoiding the generated optimization suggestions from deviating from the core needs. Furthermore, in the multi-round linear programming simulations, the actual operating resource constraints of the parking lot to be optimized are also included to ensure that the generated optimization suggestions are feasible and implementable.

[0057] Optionally, optimizing the parking lot to be optimized based on the optimization suggestions includes:

[0058] The first historical index value of the key indicator to be optimized and the second historical index value of the benchmark key indicator are collected according to the first preset historical period. A first prediction model is constructed based on the first historical index value using the least squares method, and a second prediction model is constructed based on the second historical index value using the least squares method.

[0059] Collect the third historical index value of the key index to be optimized according to the second preset historical period, input the third historical index value into the first prediction model for prediction, generate the historical benchmark prediction value, and simultaneously input the third historical index value into the second prediction model for prediction, generate the historical ideal prediction value.

[0060] Using the historical ideal predicted value and the historical benchmark predicted value as inputs, a linear regression model is used to construct a calibration model that can correct the historical benchmark predicted value, with the actual index value corresponding to the third historical index value as the output.

[0061] Obtain the current index value of the key difference index, input the current index value into the first prediction model for prediction to generate the current benchmark prediction value, and simultaneously input the current index value into the second prediction model for prediction to generate the current ideal prediction value. Input the current benchmark prediction value and the current ideal prediction value into the calibration model to obtain the corrected current benchmark prediction value.

[0062] The optimization suggestions are dynamically adjusted based on the corrected current baseline prediction values ​​to obtain dynamically adjusted optimization suggestions. The parking lot to be optimized is then optimized based on the dynamically adjusted optimization suggestions.

[0063] Optionally, dynamically adjusting the optimization recommendations based on the corrected current baseline prediction includes:

[0064] By incorporating the corrected current baseline prediction into the linear regression model, a new linear regression model is obtained, which is:

[0065]

[0066] in, This indicates a new operating index for parking lots that need optimization. This represents the corrected current baseline forecast value for the nth key difference indicator. This represents the weight of the impact of the revised current benchmark forecast of the nth differential key indicator on the operating index.

[0067] The first correlation constraint and the upper limit constraint of the index are updated based on the new linear regression model, resulting in the updated first correlation constraint and the updated upper limit constraint of the index. The updated first correlation constraint is as follows:

[0068] ;

[0069] in, This represents the forecast correction factor influenced by the future maximum daytime operating potential and the future maximum nighttime operating potential.

[0070] The updated upper limit constraint for the indicator is:

[0071] ;

[0072] in, The forecast correction factor represents the impact of the revised current baseline forecast.

[0073] The optimization suggestions are dynamically adjusted based on the updated first association constraint and the updated upper limit constraint of the index.

[0074] As described above, the optimization suggestions are dynamically adjusted based on the corrected current baseline prediction value. This allows the suggestions to adapt to future changes in the parking lot to be optimized, preventing them from becoming ineffective due to changes in actual conditions and ensuring their continued effectiveness. Furthermore, the corrected current baseline prediction value is obtained through a calibration model. This model is generated by combining historical baseline prediction values ​​obtained from a first prediction model constructed using first historical index values ​​with historical baseline prediction values ​​obtained from a second prediction model constructed using second historical index values, along with actual index values. This improves the accuracy and reliability of the resulting corrected current baseline prediction value.

[0075] Secondly, the present invention provides a parking lot operation optimization system based on multi-dimensional analysis, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the parking lot operation optimization method based on multi-dimensional analysis described in the first aspect.

[0076] The technical effects of the parking lot operation optimization system based on multi-dimensional analysis provided in the second aspect are the same as those of the parking lot operation optimization method based on multi-dimensional analysis provided in the first aspect. Attached Figure Description

[0077] Figure 1 This is a flowchart illustrating a parking lot operation optimization method based on multi-dimensional analysis provided in this embodiment;

[0078] Figure 2 This is a schematic diagram of the overall process of a parking lot operation optimization method based on multi-dimensional analysis provided in this embodiment;

[0079] Figure 3 This is a schematic diagram illustrating the process of dynamically adjusting optimization suggestions as described in this embodiment;

[0080] Figure 4 This is a schematic diagram of the structure of a parking lot operation optimization system based on multi-dimensional analysis provided in this embodiment.

[0081] [Explanation of Labels in the Attached Image]

[0082] 1. A parking lot operation optimization system based on multi-dimensional analysis;

[0083] 2. Processor;

[0084] 3. Memory. Detailed Implementation

[0085] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the present invention can be understood more clearly and thoroughly, and that the scope of the present invention can be fully conveyed to those skilled in the art.

[0086] Example 1

[0087] Please refer to Figures 1 to 3 This invention provides a parking lot operation optimization method based on multi-dimensional analysis, including the following steps:

[0088] S1. Based on the constructed multi-dimensional indicator system, collect the first multi-dimensional data of the parking lot to be optimized and the second multi-dimensional data of the parking lot to be selected, calculate the similarity between the first multi-dimensional data and the second multi-dimensional data, and take all the parking lots to be selected with the similarity exceeding the first similarity threshold as a similar parking lot cluster. The multi-dimensional indicator system includes basic indicator attributes, operational indicator attributes and environmental indicator attributes.

[0089] In this embodiment, as Figure 2 As shown, a multi-dimensional indicator system was pre-constructed, including basic indicator attributes, operational indicator attributes, and environmental indicator attributes. The basic indicator attributes include multiple sub-dimensions, such as parking lot type, business type, and parking space scale. Parking lot type is divided into two categories: on-street parking and off-street parking. Business type includes commercial districts, residential areas, and transportation hubs. The operational indicator attributes include multiple sub-dimensions, such as fee structure, operating duration, and monthly rent ratio. The fee structure includes average fee rate, charging model, and preferential policies. The environmental indicator attributes include subway environment, bus route environment, and main road environment. Based on the constructed multi-dimensional indicator system, first-dimensional data of the parking lot to be optimized and second-dimensional data of the parking lots to be selected were collected. At this point, the parking lots to be selected refer to all parking lots other than those to be optimized. The similarity between the first-dimensional data and the second-dimensional data was calculated. Parking lots to be selected with a similarity exceeding a first similarity threshold were grouped into similar parking lot clusters. The first similarity threshold was 75%, which can be adjusted according to actual conditions.

[0090] At this point, in step S1, the first multi-dimensional data includes first basic indicator data, first operational indicator data, and first environmental indicator data; the second multi-dimensional data includes second basic indicator data, second operational indicator data, and second environmental indicator data. Calculating the similarity between the first multi-dimensional data and the second multi-dimensional data, and identifying all candidate parking lots with similarity exceeding a first similarity threshold as a cluster of similar parking lots, includes:

[0091] S11. Calculate the basic similarity between the first basic indicator data and the second basic indicator data using the first similarity formula; simultaneously calculate the operational similarity between the first operational indicator data and the second operational indicator data using the second similarity formula; and calculate the environmental similarity between the first environmental indicator data and the second environmental indicator data using the third similarity formula. The first similarity formula is:

[0092] ;

[0093] in, This represents the basic similarity between parking lot i to be optimized and parking lot j to be selected. Indicates the similarity of parking lot types. Indicates the similarity of business types. Indicates the similarity in parking space size. This indicates the weights affected by the similarity of parking lot types. This indicates the weighting influenced by the similarity of business types. This indicates the weight influenced by the similarity in parking space size;

[0094] The second similarity formula is:

[0095] ;

[0096] in, This represents the operational similarity between parking lot i to be optimized and parking lot j to be selected. Indicates the similarity of rate structures. Indicates the similarity of operating duration. Indicates the similarity of monthly rent ratios. This indicates the weights affected by the similarity of the rate structure. This indicates the weighting influenced by the similarity of operating duration. This indicates the weight influenced by the similarity of monthly rent ratios;

[0097] The third similarity formula is:

[0098] ;

[0099] in, This represents the environmental similarity between parking lot i to be optimized and parking lot j to be selected. Indicating the similarity of subway environments, Indicates the similarity of bus routes. Indicates the environmental similarity of main roads. This indicates the weight influenced by the similarity of the subway environment. This indicates the weights affected by the similarity of bus routes. This indicates the weight influenced by the similarity of the main road environment;

[0100] S12. Input the basic similarity, the operational similarity, and the environmental similarity into the total similarity formula for calculation to obtain the total similarity, wherein the total similarity formula is:

[0101] ;

[0102] in, This represents the total similarity between parking lot i to be optimized and parking lot j to be selected. This represents the weights that influence the basic similarity. This indicates the weight that influences the similarity of operations. This represents the weight that influences environmental similarity;

[0103] S13. All parking lots to be selected with a total similarity exceeding the first similarity threshold are taken as a similar parking lot cluster.

[0104] In this embodiment, as Figure 2 As shown, the first and second multi-dimensional data correspond to the multi-dimensional indicator system. Therefore, the first multi-dimensional data includes the first basic indicator data, the first operational indicator data, and the first environmental indicator data; the second multi-dimensional data includes the second basic indicator data, the second operational indicator data, and the second environmental indicator data. Different formulas are used to calculate the corresponding similarity for data of different dimensions. According to the first, second, and third similarity formulas, the basic similarity, operational similarity, and environmental similarity are all obtained by weighted summation of the similarities of different types of sub-dimensions. Furthermore, corresponding weights are set for the similarities of different types of sub-dimensions, meaning the weights dynamically change with the similarity of the corresponding sub-dimensions. Specific weight assignments can be matched according to a pre-defined mapping table between sub-dimension similarity and weights.

[0105] In the first similarity formula, the parking lot type similarity is calculated as follows: when the parking lot type of the parking lot to be selected is the same as that of the parking lot to be optimized, the parking lot type similarity is directly assigned a value of 1; otherwise, it is directly assigned a value of 0. The business type similarity is calculated as follows: when the business type of the parking lot to be selected is exactly the same as that of the parking lot to be optimized, the business type similarity is directly assigned a value of 1; otherwise, it is directly assigned a value of 0. The parking space size similarity is calculated as follows: Parking space size similarity = 1 - |Number of parking spaces in the parking lot to be optimized - Number of parking spaces in the parking lot to be selected| / max(Number of parking spaces in the parking lot to be optimized, Number of parking spaces in the parking lot to be selected).

[0106] In the second similarity formula, the rate structure similarity is calculated as follows: Rate structure similarity = Average rate similarity + Charging model similarity + Preferential policy similarity;

[0107] Average rate similarity = (daytime rate similarity + nighttime rate similarity + daily cap similarity + monthly rent similarity) / 4;

[0108] Daytime rate similarity = (min(daytime hourly rate) / max(daytime hourly rate);

[0109] Nighttime rate similarity = min(nighttime hourly rate) / max(nighttime hourly rate);

[0110] Daily cap similarity = min(daily cap cost) / max(daily cap cost);

[0111] Monthly rental similarity = min(monthly rental cost) / max(monthly rental cost);

[0112] Charging pattern similarity = max(0, 1 - charging pattern difference × first penalty coefficient) = max(0, 1 - |charging pattern mapping value of the parking lot to be selected - charging pattern mapping value of the parking lot to be optimized| × penalty coefficient).

[0113] The daytime period refers to 8:00-18:00, and the nighttime period refers to 18:00-8:00 the next day. The charging mode mapping values ​​are: hourly billing → 1, tiered billing → 2; segmented billing → 3, capped billing → 4, and mixed billing → 5. The first penalty coefficient is 0.2.

[0114] Similarity of preferential policies = |Preferential policies of parking lots to be selected ∩ Preferential policies of parking lots to be optimized| / |Preferential policies of parking lots to be selected ∪ Preferential policies of parking lots to be optimized|;

[0115] The preferential policies include: member discounts, daytime discounts, nighttime discounts, points redemption, and monthly subscription discounts.

[0116] The calculation method for operation duration similarity is: Operation duration similarity = 1 - |Daily operation duration of the parking lot to be selected - Daily operation duration of the parking lot to be optimized| / Standard value of daily operation duration;

[0117] The calculation method for monthly rental ratio similarity is: monthly rental ratio similarity = 1 - |monthly rental ratio of the parking lot to be selected - monthly rental ratio of the parking lot to be optimized|;

[0118] In the third similarity formula, the subway environment similarity is calculated as follows: Subway environment similarity = 1 - |the shortest distance between the parking lot to be selected and the subway - the shortest distance between the parking lot to be optimized and the subway| / 1000;

[0119] The similarity of bus routes is calculated as follows: Bus route similarity = 1 - |Number of bus routes in parking lots to be selected - Number of bus routes in parking lots to be optimized| / max(Number of bus routes in parking lots to be selected, Number of bus routes in parking lots to be optimized).

[0120] The calculation method for the environmental similarity of the main road is: Environmental similarity of the main road = 1 - |the shortest distance between the parking lot to be selected and the main road - the shortest distance between the parking lot to be optimized and the main road| / 1000.

[0121] According to the overall similarity formula, the overall similarity is obtained by weighted summation of basic similarity, operational similarity, and environmental similarity. While different weights are assigned to different dimensions of similarity in the overall similarity formula, unlike the formulas for calculating similarity in different dimensions (i.e., the first, second, and third similarity formulas), the weights in the overall similarity formula are pre-set and do not affect the similarity of different dimensions. Instead, they influence the similarity of different dimensions inversely. These weights are set according to specific business needs. Initially, the weight affecting basic similarity is set to 0.25, the weight affecting operational similarity to 0.6, and the weight affecting environmental similarity to 0.15. If subsequent business needs lean towards environmental indicators, the weight of environmental similarity can be increased; similarly, if subsequent business needs lean towards basic indicators, the weight of basic similarity can be increased. The specific settings are dynamically adjusted based on subsequent business needs.

[0122] S2. Obtain the actual operating value and maximum operating potential value of each parking lot to be selected in the similar parking lot cluster. Calculate the corresponding operating index based on the actual operating value and maximum operating potential value of each parking lot to be selected. Use the parking lot with the highest operating index as the benchmark parking lot of the parking lot to be optimized.

[0123] In this embodiment, as Figure 2 As shown, the parking lot with the highest operating index is selected from the similar parking lot clusters obtained in step S1 as the benchmark parking lot to be optimized.

[0124] At this point, step S2, which involves obtaining the actual operating value and maximum operating potential value of each parking lot to be selected in the similar parking lot cluster, and calculating the corresponding operating index based on the actual operating value and maximum operating potential value of each parking lot to be selected, includes:

[0125] S21. Calculate the parking space turnover rate and parking space utilization rate of each parking lot to be selected in the similar parking lot cluster during the daytime and nighttime periods respectively, to obtain the daytime turnover rate, nighttime turnover rate, daytime utilization rate and nighttime utilization rate of each parking lot to be selected, and calculate the daytime revenue ratio of each parking lot to be selected based on the daytime turnover rate and daytime utilization rate of each parking lot to be selected, and at the same time calculate the nighttime revenue ratio of each parking lot to be selected based on the nighttime turnover rate and nighttime utilization rate of each parking lot to be selected.

[0126] S22. Calculate the actual daytime operating value of each parking lot based on its daytime turnover rate, daytime utilization rate, and daytime revenue ratio. At the same time, calculate the actual nighttime operating value of each parking lot based on its nighttime turnover rate, nighttime utilization rate, and nighttime revenue ratio.

[0127] S23. Obtain the daytime operating hours and average daytime parking time corresponding to the daytime utilization rate of each candidate parking lot. At the same time, obtain the nighttime operating hours and average nighttime parking time corresponding to the nighttime utilization rate of each candidate parking lot. Calculate the potential daytime turnover rate of each candidate parking lot based on the daytime utilization rate, daytime operating hours and average daytime parking time. Calculate the potential nighttime turnover rate of each candidate parking lot based on the nighttime utilization rate, nighttime operating hours and average nighttime parking time.

[0128] S24. Select the maximum utilization rate of each time period from the daytime utilization rate of each parking lot to be selected as the potential value of daytime utilization rate, and at the same time select the maximum utilization rate of each time period from the nighttime utilization rate of each parking lot to be selected as the potential value of nighttime utilization rate.

[0129] S25. Calculate the maximum daytime operating potential value based on the daytime turnover rate potential value and the daytime utilization rate potential value of each parking lot to be selected. At the same time, calculate the maximum nighttime operating potential value based on the nighttime turnover rate potential value and the nighttime utilization rate potential value of each parking lot to be selected. Use the ratio of the actual daytime operating value of each parking lot to the maximum daytime operating potential value as the daytime operating index of each parking lot to be selected. Use the ratio of the actual nighttime operating value of each parking lot to the maximum nighttime operating potential value as the nighttime operating index of each parking lot to be selected.

[0130] S26. Obtain the final operating index of each parking lot to be selected based on its daytime operating index and nighttime operating index.

[0131] In this embodiment, as Figure 2As shown, the parking space turnover rate and parking space utilization rate of each candidate parking lot in the similar parking lot cluster are calculated during the daytime and nighttime periods, respectively, to obtain the daytime turnover rate, nighttime turnover rate, daytime utilization rate, and nighttime utilization rate for each candidate parking lot. The daytime period refers to 8:00-18:00, and the nighttime period refers to 18:00-8:00 the next day. The parking space turnover rate refers to the average number of times each parking space is used by different vehicles within a specific time period, i.e., parking space turnover rate = total number of vehicles entering the parking lot during the time period / total number of parking spaces. The parking space utilization rate refers to the average parking space occupancy at each hour within a specific time period, i.e., parking space utilization rate = number of parking spaces in use at each hour / total number of parking spaces = (number of vehicles in the parking lot at the previous hour = number of vehicles leaving the parking lot at the previous hour + number of vehicles entering the parking lot at the current hour) / total number of parking spaces. The daytime revenue ratio for each parking lot is calculated based on its daytime turnover rate and daytime utilization rate. Similarly, the nighttime revenue ratio for each parking lot is calculated based on its nighttime turnover rate and nighttime utilization rate.

[0132] The actual daytime operating value of each parking lot is calculated based on its daytime turnover rate, daytime utilization rate, and daytime revenue ratio. Specifically, the actual daytime operating value = (daytime turnover rate × daytime utilization rate × daytime revenue ratio)^(1 / 3). At the same time, the actual nighttime operating value of each parking lot is calculated based on its nighttime turnover rate, nighttime utilization rate, and nighttime revenue ratio. Specifically, the actual nighttime operating value = (nighttime turnover rate × nighttime utilization rate × nighttime revenue ratio)^(1 / 3).

[0133] Obtain the daytime operating hours and average daytime parking time corresponding to the daytime utilization rate of each candidate parking lot. Calculate the daytime turnover potential value for each candidate parking lot based on the daytime utilization rate, daytime operating hours, and average daytime parking time. Specifically, the daytime turnover potential value = daytime utilization rate × daytime operating hours ÷ average daytime parking time. Similarly, calculate the nighttime turnover potential value. Select the maximum utilization rate for a specific time period from the daytime utilization rates corresponding to each candidate parking lot as the daytime utilization potential value. Simultaneously, select the maximum utilization rate for a specific time period from the nighttime utilization rates as the nighttime utilization potential value. Calculate the maximum daytime operating potential value based on the daytime turnover potential value and daytime utilization potential value for each candidate parking lot. Specifically: Maximum daytime operating potential value = (Daytime turnover potential value × Daytime utilization potential value)^(1 / 3). Simultaneously, calculate the maximum nighttime operating potential value based on the nighttime turnover potential value and nighttime utilization potential value for each candidate parking lot. Maximum nighttime operating potential value = (Nighttime turnover potential value × Nighttime utilization potential value)^(1 / 3). The ratio of the actual daytime operating value to the maximum daytime operating potential value of each candidate parking lot is used as the daytime operating index of each candidate parking lot, where the daytime operating index = actual daytime operating value / maximum daytime operating potential value; the ratio of the actual nighttime operating value to the maximum nighttime operating potential value of each candidate parking lot is used as the nighttime operating index of each candidate parking lot, where the nighttime operating index = actual nighttime operating value / maximum nighttime operating potential value.

[0134] The final operating index for each parking lot is calculated based on its daytime and nighttime operating indices. The final operating index is calculated as follows: (Daytime operating index ^ Daytime weight) × (Nighttime operating index ^ Nighttime weight). The daytime weight is set based on the ratio of traffic flow to parking time during specific daytime periods, and the nighttime weight is set based on the ratio of traffic flow to parking time during specific nighttime periods.

[0135] S3. Based on the constructed key indicator system, collect the benchmark key indicators of the benchmark parking lot and the key indicators to be optimized of the parking lot to be optimized, compare the key indicators to be optimized with the benchmark key indicators, and generate the difference key indicators.

[0136] In this embodiment, as Figure 2 Therefore, a key indicator system was pre-constructed, which includes: business type, number of parking spaces, total traffic flow, turnover rate, average parking space utilization rate, average dwell time, total revenue, average parking space revenue, revenue per unit time, revenue structure ratio, fee structure, and operating costs. Based on the constructed key indicator system, benchmark key indicators of benchmark parking lots and key indicators to be optimized of parking lots to be optimized were collected and compared to generate difference key indicators.

[0137] S4. Based on the key difference indicators and the actual operational resource constraints of the parking lot to be optimized, generate feasible optimization suggestions, and optimize the parking lot to be optimized according to the optimization suggestions.

[0138] At this point, step S4, which involves generating feasible optimization suggestions based on the key difference indicators and the actual operational resource constraints of the parking lot to be optimized, and then optimizing the parking lot according to the optimization suggestions, includes:

[0139] S41. Based on the aforementioned key difference indicators, construct a linear regression model related to the operating index of the parking lot to be optimized, wherein the linear regression model is:

[0140]

[0141] in, This indicates the operational index of parking lots that need optimization. This represents the current value of the nth key difference indicator. This represents the weight of the nth key difference indicator on the operating index. Represents the residual. Indicates the total number of key indicators of difference;

[0142] S42. Obtain the operating index of the benchmark parking lot, and simultaneously calculate the current index difference between the operating index of the benchmark parking lot and the operating index of the parking lot to be optimized. Based on the current index difference and the linear regression model, construct a first correlation constraint. The first correlation constraint is:

[0143] ;

[0144] in, This represents the current value of the key indicator for the nth benchmark parking lot. This indicates the operating index of benchmark parking lots;

[0145] S43. Using the minimum optimization cost of the parking lot to be optimized as the objective function, under the first association constraint and the actual operating resource constraint of the parking lot to be optimized, a feasible optimization suggestion is generated through multiple rounds of linear programming simulation. The optimization of the parking lot to be optimized is realized according to the optimization suggestion, wherein the optimization suggestion includes the optimization target value of each difference key indicator.

[0146] The objective function is:

[0147] ;

[0148] ;

[0149] in, Let represent the objective function of the parking lot to be optimized. This represents the current value of the nth key difference indicator. This represents the current value of the key indicator for the nth benchmark parking lot. This represents the optimization cost of the nth key difference indicator. The cost of investing in optimizing the nth key difference indicator. The time cost of optimizing the nth key difference indicator;

[0150] The actual operational resource constraints include: indicator upper limit constraints, second correlation constraints, and resource upper limit constraints, wherein the indicator upper limit constraints are:

[0151] ;

[0152] in, This represents the upper limit of improvement for the nth key difference indicator;

[0153] The second association constraint is:

[0154] ;

[0155] in, This represents the lower limit threshold of the association between the nth key difference indicator and the mth key difference indicator. This represents the correlation coefficient between the nth difference key indicator and the nth benchmark key indicator. This represents the correlation coefficient between the m-th difference key indicator and the m-th benchmark key indicator. This represents the upper limit threshold for the association between the nth differential key indicator and the mth differential key indicator;

[0156] The resource upper limit constraint is:

[0157] ;

[0158] in, This indicates the maximum available investment in the parking lot to be optimized. This indicates the maximum available time for the parking lot to be optimized.

[0159] In this embodiment, as Figure 2As shown, a linear regression model is constructed based on key difference indicators (KPIs) to correlate with the operating index of the parking lot to be optimized. The aim is to quantify the correlation between KPIs and the operating index, providing a basis for constructing the first correlation constraint. The operating index of the benchmark parking lot is obtained, and the current index difference between the benchmark parking lot's operating index and the operating index of the parking lot to be optimized is calculated. That is, the incremental operating index required for the parking lot to be optimized to reach the benchmark parking lot's operating index is calculated. According to the first correlation constraint, it ensures that the improvement effect of the KPIs of the parking lot to be optimized accurately points to the incremental operating index, achieving a strong correlation between the KPIs and the improvement of the operating index of the parking lot to be optimized.

[0160] Using the minimum optimization cost of the parking lot to be optimized as the objective function, and under the constraints of the first correlation constraint and the actual operating resources of the parking lot, multiple rounds of linear programming simulation are used to generate feasible optimization suggestions containing the optimization target value for each key difference indicator. The optimization of the parking lot to be optimized is then achieved based on these suggestions. The optimization cost in the objective function includes not only the financial cost but also the time cost. The actual operating resource constraints of the parking lot to be optimized include indicator upper limit constraints, a second correlation constraint, and a resource upper limit constraint. The indicator upper limit constraint represents the upper limit of improvement for each key difference indicator; the resource upper limit constraint represents the maximum available financial investment and the maximum available time investment; and the second correlation constraint mainly refers to the correlation constraints between different key difference indicators.

[0161] At this point, the optimization of the parking lot to be optimized based on the optimization suggestions in step S43 includes:

[0162] S431. Collect the first historical index value of the key indicator to be optimized and the second historical index value of the benchmark key indicator according to the first preset historical period. Construct a first prediction model based on the first historical index value using the least squares method, and simultaneously construct a second prediction model based on the second historical index value using the least squares method.

[0163] S432. Collect the third historical index value of the key index to be optimized according to the second preset historical period, input the third historical index value into the first prediction model for prediction, generate the historical benchmark prediction value, and simultaneously input the third historical index value into the second prediction model for prediction, generate the historical ideal prediction value.

[0164] S433. Using the historical ideal predicted value and the historical benchmark predicted value as inputs, a linear regression model is used to construct a calibration model that can correct the historical benchmark predicted value, with the actual index value corresponding to the third historical index value as the output.

[0165] S434. Obtain the current index value of the key difference index, input the current index value into the first prediction model for prediction, generate the current benchmark prediction value, and simultaneously input the current index value into the second prediction model for prediction, generate the current ideal prediction value, and input the current benchmark prediction value and the current ideal prediction value into the calibration model to obtain the corrected current benchmark prediction value.

[0166] S435. The optimization suggestions are dynamically adjusted based on the corrected current baseline prediction value to obtain dynamically adjusted optimization suggestions, and the parking lot to be optimized is optimized according to the dynamically adjusted optimization suggestions.

[0167] In this embodiment, as Figure 3 As shown, the optimization suggestions incorporate dynamic prediction for the parking lot to be optimized. Historical values ​​of the key indicators to be optimized and the benchmark key indicators are collected according to a first preset historical period (the past 24 months). A first prediction model is constructed using the least squares method based on the first historical value, capturing the operational patterns of the parking lot to be optimized and achieving prediction for it. Simultaneously, a second prediction model is constructed using the least squares method based on the second historical value, capturing the operational patterns of the benchmark parking lot and achieving prediction for it. A third historical value of the key indicators to be optimized is collected based on the second preset historical period (shorter than the first, e.g., the past 6 months). The third historical value is input into the first prediction model (the prediction model for the parking lot to be optimized) to generate historical baseline prediction values. Simultaneously, the third historical value is input into the second prediction model (the prediction model for the benchmark parking lot) to generate ideal prediction values ​​from the benchmark perspective, i.e., historical ideal prediction values.

[0168] Using historical ideal predicted values ​​and historical benchmark predicted values ​​as inputs, a linear regression model is employed, with the actual index value corresponding to the third historical index value as the output, to construct a calibration model capable of correcting the historical benchmark predicted values. In this embodiment, the constructed calibration model undergoes temporal cross-validation, progressing sequentially over time to ensure dynamic updates. The mean absolute percentage error (MAE) and root mean square error (RMSE) are used to evaluate the effectiveness of each calibration model. If the evaluation results do not meet expectations, the calibration model is retrained until the desired effect is achieved.

[0169] The current values ​​of key performance indicators (KPIs) are obtained and input into a first prediction model to generate a current baseline prediction value. Simultaneously, the current KPIs are input into a second prediction model to generate a current ideal prediction value. The obtained current baseline and ideal prediction values ​​are then input into a calibration model to correct the current baseline prediction value, resulting in a corrected current baseline prediction value. In other words, the future prediction value of the parking lot to be optimized is obtained, i.e., the corrected current baseline prediction value. Based on the corrected current baseline prediction value, the optimization suggestions are dynamically adjusted to ensure their sustainability, resulting in dynamically adjusted optimization suggestions. The parking lot to be optimized is then optimized sequentially based on these dynamically adjusted suggestions.

[0170] At this point, the dynamic adjustment of the optimization suggestion based on the corrected current baseline prediction value in step S435 includes:

[0171] S4351. Introducing the corrected current baseline prediction into the linear regression model yields a new linear regression model, which is:

[0172]

[0173] in, This indicates a new operating index for parking lots that need optimization. This represents the corrected current baseline forecast value for the nth key difference indicator. This represents the weight of the impact of the revised current benchmark forecast of the nth differential key indicator on the operating index.

[0174] S4352. Based on the new linear regression model, update the first correlation constraint and the upper limit constraint of the index to obtain the updated first correlation constraint and the updated upper limit constraint of the index. The updated first correlation constraint is:

[0175] ;

[0176] in, This represents the forecast correction factor influenced by the future maximum daytime operating potential and the future maximum nighttime operating potential.

[0177] The updated upper limit constraint for the indicator is:

[0178] ;

[0179] in, The forecast correction factor represents the impact of the revised current baseline forecast.

[0180] S4353. The optimization suggestions are dynamically adjusted based on the updated first association constraint and the updated upper limit constraint of the index.

[0181] In this embodiment, as Figure 3 As shown, when dynamically adjusting the optimization suggestions, the corrected current baseline forecast value is introduced into the linear regression model. This corrected current baseline forecast value is incorporated as an independent variable, clarifying the relationship between the key difference indicators, future trends, and the operating index in the new linear regression model. Based on the new linear regression model, the first correlation constraint and the indicator upper limit constraint are updated. The updated first correlation constraint not only incorporates the corrected current baseline forecast value of the key difference indicators and their impact weight on the operating index, but also introduces a prediction correction coefficient influenced by the future maximum daytime operating potential value and the future maximum nighttime operating potential value. The prediction correction coefficient increases as the sum of the obtained future maximum daytime operating potential value and the future maximum nighttime operating potential value increases, with an initial value of 1.0. The updated first correlation constraint ensures that the optimization target value of the key difference indicators not only matches the current catch-up requirements of the key difference indicators but also takes into account the improvement potential brought by future trends. The updated upper limit constraint introduces a prediction correction coefficient affected by the revised current benchmark prediction. When the revised current benchmark is close to the current understood prediction, i.e., the difference between the two is less than 0.2, y≥1, otherwise y<1, where the value of y ranges from [0.8, 1.2].

[0182] Example 2

[0183] Please refer to Figure 4 The present invention provides a parking lot operation optimization system 1 based on multi-dimensional analysis, including a memory 3, a processor 2, and a computer program stored on the memory 3 and run on the processor 2. When the processor 2 executes the computer program, it implements the steps in Embodiment 1.

[0184] Since the systems / devices described in the above embodiments of the present invention are systems / devices used to implement the methods of the above embodiments of the present invention, those skilled in the art can understand the specific structure and modifications of the systems / devices based on the methods described in the above embodiments of the present invention, and therefore will not be repeated here. All systems / devices used in the methods of the above embodiments of the present invention fall within the scope of protection of the present invention.

[0185] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0186] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions.

[0187] It should be noted that any reference numerals placed between parentheses in the claims should not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In claims that enumerate several means, several of these means may be embodied by the same hardware. The use of the terms first, second, third, etc., is merely for convenience of expression and does not indicate any order. These terms can be understood as part of the component names.

[0188] Furthermore, it should be noted that in the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Furthermore, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0189] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the claims should be interpreted to include both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0190] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, then this invention should also include these modifications and variations.

Claims

1. A parking lot operation optimization method based on multi-dimensional analysis, characterized in that, include: Based on the constructed multi-dimensional indicator system, the first multi-dimensional data of the parking lot to be optimized and the second multi-dimensional data of the parking lot to be selected are collected. The similarity between the first multi-dimensional data and the second multi-dimensional data is calculated. All parking lots to be selected with a similarity exceeding a first similarity threshold are regarded as similar parking lot clusters. The multi-dimensional indicator system includes basic indicator attributes, operational indicator attributes and environmental indicator attributes. Obtain the actual operating value and maximum operating potential value of each parking lot to be selected in the similar parking lot cluster. Calculate the corresponding operating index based on the actual operating value and maximum operating potential value of each parking lot to be selected. Use the parking lot with the highest operating index as the benchmark parking lot for the parking lot to be optimized. Based on the constructed key indicator system, benchmark key indicators of the benchmark parking lot and key indicators to be optimized of the parking lot to be optimized are collected. The key indicators to be optimized are compared with the benchmark key indicators to generate difference key indicators. Based on the aforementioned key difference indicators and the actual operational resource constraints of the parking lot to be optimized, feasible optimization suggestions are generated. The optimization of the parking lot to be optimized is achieved according to the optimization suggestions, wherein the optimization suggestions include the optimization target value of each key difference indicator. The step of obtaining the actual operating value and maximum operating potential value of each candidate parking lot in the similar parking lot cluster, and calculating the corresponding operating index based on the actual operating value and maximum operating potential value of each candidate parking lot includes: The parking space turnover rate and parking space utilization rate of each candidate parking lot in the similar parking lot cluster are calculated during the daytime and nighttime periods, respectively, to obtain the daytime turnover rate, nighttime turnover rate, daytime utilization rate and nighttime utilization rate of each candidate parking lot. The daytime revenue ratio of each candidate parking lot is calculated based on the daytime turnover rate and daytime utilization rate of each candidate parking lot. At the same time, the nighttime revenue ratio of each candidate parking lot is calculated based on the nighttime turnover rate and nighttime utilization rate of each candidate parking lot. The actual daytime operating value of each parking lot is calculated based on its daytime turnover rate, daytime utilization rate, and daytime revenue ratio. At the same time, the actual nighttime operating value of each parking lot is calculated based on its nighttime turnover rate, nighttime utilization rate, and nighttime revenue ratio. Obtain the daytime operating hours and average daytime parking time corresponding to the daytime utilization rate of each candidate parking lot. At the same time, obtain the nighttime operating hours and average nighttime parking time corresponding to the nighttime utilization rate of each candidate parking lot. Calculate the potential daytime turnover rate of each candidate parking lot based on its daytime utilization rate, daytime operating hours, and average daytime parking time. Calculate the potential nighttime turnover rate of each candidate parking lot based on its nighttime utilization rate, nighttime operating hours, and average nighttime parking time. The maximum utilization rate during the daytime period is selected from the daytime utilization rate of each parking lot to be selected as the potential value of daytime utilization. At the same time, the maximum utilization rate during the nighttime period is selected from the nighttime utilization rate of each parking lot to be selected as the potential value of nighttime utilization. The maximum daytime operating potential is calculated based on the daytime turnover rate potential and daytime utilization rate potential of each parking lot to be selected. At the same time, the maximum nighttime operating potential is calculated based on the nighttime turnover rate potential and nighttime utilization rate potential of each parking lot to be selected. The ratio of the actual daytime operating value of each parking lot to the maximum daytime operating potential is used as the daytime operating index of each parking lot to be selected. The ratio of the actual nighttime operating value of each parking lot to the maximum nighttime operating potential is used as the nighttime operating index of each parking lot to be selected. The final operating index for each parking lot to be selected is obtained based on its daytime and nighttime operating indices. The key indicators include: business type, number of parking spaces, total traffic flow, turnover rate, average parking space utilization rate, average dwell time, total revenue, average parking space revenue, revenue per unit time, revenue structure ratio, fee structure, and operating costs.

2. The parking lot operation optimization method based on multi-dimensional analysis as described in claim 1, characterized in that, The first multi-dimensional data includes first basic indicator data, first operational indicator data, and first environmental indicator data; the second multi-dimensional data includes second basic indicator data, second operational indicator data, and second environmental indicator data; calculating the similarity between the first multi-dimensional data and the second multi-dimensional data, and identifying all candidate parking lots with similarity exceeding a first similarity threshold as a cluster of similar parking lots, includes: The basic similarity between the first basic indicator data and the second basic indicator data is calculated using a first similarity formula. Simultaneously, the operational similarity between the first operational indicator data and the second operational indicator data is calculated using a second similarity formula. Finally, the environmental similarity between the first environmental indicator data and the second environmental indicator data is calculated using a third similarity formula. The first similarity formula is as follows: ； in, This represents the basic similarity between parking lot i to be optimized and parking lot j to be selected. Indicates the similarity of parking lot types. Indicates the similarity of business types. Indicates the similarity in parking space size. This indicates the weights affected by the similarity of parking lot types. This indicates the weighting influenced by the similarity of business types. This indicates the weight influenced by the similarity in parking space size; The second similarity formula is: ； in, This represents the operational similarity between parking lot i to be optimized and parking lot j to be selected. Indicates the similarity of rate structures. Indicates the similarity of operating duration. Indicates the similarity of monthly rent ratios. This indicates the weights affected by the similarity of the rate structure. This indicates the weighting influenced by the similarity of operating duration. This indicates the weight influenced by the similarity of monthly rent ratios; The third similarity formula is: ； in, This represents the environmental similarity between parking lot i to be optimized and parking lot j to be selected. Indicates the similarity of subway environments. Indicates the similarity of bus routes. Indicates the similarity of the main road environment. This indicates the weights influenced by the similarity of the subway environment. This indicates the weights affected by the similarity of bus routes. This indicates the weight influenced by the similarity of the main road environment; The basic similarity, the operational similarity, and the environmental similarity are input into the total similarity formula for calculation to obtain the total similarity, wherein the total similarity formula is: ； in, This represents the total similarity between parking lot i to be optimized and parking lot j to be selected. This represents the weights that influence the basic similarity. This indicates the weight that influences the similarity of operations. This represents the weight that influences environmental similarity; All parking lots to be selected whose total similarity exceeds the first similarity threshold are considered as a similar parking lot cluster.

3. The parking lot operation optimization method based on multi-dimensional analysis as described in claim 1, characterized in that, The process of generating feasible optimization suggestions based on the key difference indicators and the actual operational resource constraints of the parking lot to be optimized, and then optimizing the parking lot to be optimized according to the optimization suggestions, includes: Based on the aforementioned key difference indicators, a linear regression model is constructed that relates to the operating index of the parking lot to be optimized. The linear regression model is as follows: in, This indicates the operational index of parking lots that need optimization. This represents the current value of the nth key difference indicator. This represents the weight of the nth key difference indicator on the operating index. Represents the residual. Indicates the total number of key indicators of difference; Obtain the operating index of the benchmark parking lot, and simultaneously calculate the current index difference between the operating index of the benchmark parking lot and the operating index of the parking lot to be optimized. Based on the current index difference and the linear regression model, construct a first correlation constraint, which is: ； in, This represents the current value of the key indicator for the nth benchmark parking lot. This indicates the operating index of benchmark parking lots; Using the minimum optimization cost of the parking lot to be optimized as the objective function, and under the first association constraint and the actual operating resource constraint of the parking lot to be optimized, a multi-round linear programming simulation is used to generate an executable optimization suggestion. The optimization of the parking lot to be optimized is realized according to the optimization suggestion, wherein the optimization suggestion includes the optimization target value of each difference key indicator. The objective function is: ； ; in, Let represent the objective function of the parking lot to be optimized. This represents the current value of the nth key difference indicator. This represents the current value of the key indicator for the nth benchmark parking lot. This represents the optimization cost of the nth key performance indicator (KPI). The cost of investing in optimizing the nth key difference indicator. The time cost of optimizing the nth key difference indicator; The actual operational resource constraints include: indicator upper limit constraints, second correlation constraints, and resource upper limit constraints, wherein the indicator upper limit constraints are: ； in, This represents the upper limit of improvement for the nth key difference indicator; The second association constraint is: ； in, This represents the lower limit threshold of the association between the nth key difference indicator and the mth key difference indicator. This represents the correlation coefficient between the nth difference key indicator and the nth benchmark key indicator. This represents the correlation coefficient between the m-th difference key indicator and the m-th benchmark key indicator. This represents the upper limit threshold for the association between the nth differential key indicator and the mth differential key indicator; The resource upper limit constraint is: ； in, This indicates the maximum available investment in the parking lot to be optimized. This indicates the maximum available time for the parking lot to be optimized.

4. The parking lot operation optimization method based on multi-dimensional analysis as described in claim 3, characterized in that, The optimization of the parking lot to be optimized based on the optimization suggestions includes: The first historical index value of the key indicator to be optimized and the second historical index value of the benchmark key indicator are collected according to the first preset historical period. A first prediction model is constructed based on the first historical index value using the least squares method, and a second prediction model is constructed based on the second historical index value using the least squares method. Collect the third historical index value of the key index to be optimized according to the second preset historical period, input the third historical index value into the first prediction model for prediction, generate the historical benchmark prediction value, and simultaneously input the third historical index value into the second prediction model for prediction, generate the historical ideal prediction value. Using the historical ideal predicted value and the historical benchmark predicted value as inputs, a linear regression model is used to construct a calibration model that can correct the historical benchmark predicted value, with the actual index value corresponding to the third historical index value as the output. Obtain the current index value of the key difference index, input the current index value into the first prediction model for prediction to generate the current benchmark prediction value, and simultaneously input the current index value into the second prediction model for prediction to generate the current ideal prediction value. Input the current benchmark prediction value and the current ideal prediction value into the calibration model to obtain the corrected current benchmark prediction value. The optimization suggestions are dynamically adjusted based on the corrected current baseline prediction values ​​to obtain dynamically adjusted optimization suggestions. The parking lot to be optimized is then optimized based on the dynamically adjusted optimization suggestions.

5. The parking lot operation optimization method based on multi-dimensional analysis as described in claim 4, characterized in that, The dynamic adjustment of the optimization suggestions based on the corrected current baseline prediction includes: By incorporating the corrected current baseline prediction into the linear regression model, a new linear regression model is obtained, which is: in, This indicates a new operating index for parking lots that need optimization. This represents the corrected current baseline forecast value for the nth key difference indicator. This represents the weight of the impact of the revised current benchmark forecast of the nth differential key indicator on the operating index. The first correlation constraint and the upper limit constraint of the index are updated based on the new linear regression model, resulting in the updated first correlation constraint and the updated upper limit constraint of the index. The updated first correlation constraint is as follows: ； in, This represents the forecast correction factor influenced by the future maximum daytime operating potential and the future maximum nighttime operating potential. The updated upper limit constraint for the indicator is: ； in, The forecast correction factor represents the impact of the revised current baseline forecast. The optimization suggestions are dynamically adjusted based on the updated first association constraint and the updated upper limit constraint of the index.

6. A parking lot operation optimization system based on multi-dimensional analysis, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 5.

Images