Outdoor space public facility layout optimization method based on big data analysis

Abstract

The invention discloses an outdoor space public facility layout optimization method based on big data analysis, and belongs to the field of outdoor space public facility layout optimization, and the method comprises the steps: predicting the number of participants in a data set based on demands, and combining the illumination intensity, wind direction, temperature and rainfall in a comprehensive data set, calculating a meteorological suitability score of each region, and generating a preliminary facility layout data set including candidate position points and meteorological suitability scores thereof; on the basis of the adjusted candidate position data set, in combination with the average decibel value of each region, spatial analysis is executed, position points with the average decibel value lower than 60 decibels are screened, and an optimized position data set is generated and comprises the position points, a meteorological suitability score and a noise level; and based on the optimized position data set, in combination with the facility coverage range and the construction and maintenance cost, calculating the facility number and type of each region, and generating a facility layout planning data set including a facility configuration list of each region.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of outdoor space public facility layout optimization, and particularly relates to an outdoor space public facility layout optimization method based on big data analysis. BACKGROUND

[0002] Campus outdoor space public facility layout optimization is a crucial research field, directly related to student quality of life, campus resource utilization efficiency, and environmental harmony development. With the expansion of university scale and the diversification of student needs, reasonable allocation of outdoor facilities has become a key to improving campus management level and user satisfaction. However, current related methods often have significant limitations, mainly in the lack of comprehensive consideration of dynamic demand and complex environment. Many solutions rely more on static planning or experience judgment, making it difficult to adapt to changes in student behavior, seasonal fluctuations, and changing environmental conditions, leading to a mismatch between facility layout and actual demand, resource waste, and inconvenience.

[0003] Under this background, the core challenges in this field gradually emerge. First, student use of campus outdoor space has high spatiotemporal dynamics, with significant differences in demand intensity at different times and in different areas. Existing technologies often struggle to accurately predict these changes. The root of this prediction difficulty lies in the lack of integration and analysis capabilities of multi-source information, such as fragmented data on weather conditions, course schedules, and activity schedules, making demand assessment lack comprehensiveness and accuracy. Further, this lack of information integration directly affects the scientificity of facility layout, especially in the face of environmental factors such as light, wind direction, and noise, making it difficult to effectively match facility location and environmental conditions, resulting in poor performance of layout schemes in actual application. SUMMARY

[0004] To solve the above technical problems, the application provides an outdoor space public facility layout optimization method based on big data analysis, comprising:

[0005] Obtaining campus multi-source data and generating a comprehensive data set;

[0006] Extracting behavior patterns from the comprehensive data set and generating a behavior pattern data set;

[0007] Predicting demand from the behavior pattern data set and generating a demand prediction data set;

[0008] Calculating weather suitability scores from the demand prediction data set and weather data and generating a preliminary facility layout data set;

[0009] Performing weather condition screening based on the preliminary facility layout data set and generating an adjusted candidate location data set;

[0010] performing spatial analysis on the adjusted candidate location dataset in combination with noise level and generating an optimized location dataset;

[0011] performing spatial analysis on the adjusted candidate location dataset in combination with noise level and generating an optimized location dataset;

[0012] performing spatial analysis on the adjusted candidate location dataset in combination with noise level and generating an optimized location dataset;

[0013] performing spatial analysis on the adjusted candidate location dataset in combination with noise level and generating an optimized location dataset.

[0014] Optionally, the obtaining of the campus multi-source data and the generation of the comprehensive dataset comprises:

[0015] collecting course schedule, activity schedule and weather data from the campus management system and the weather interface;

[0016] unifying the time stamp of the data into ISO format and performing numerical normalization;

[0017] adopting a proximity algorithm to fill in missing values and storing in a unified database;

[0018] clustering and merging the data in the database according to time and place to generate the comprehensive dataset.

[0019] Optionally, the extraction of the behavior pattern from the comprehensive dataset and the generation of the behavior pattern dataset comprises:

[0020] dividing the number of participants and the activity type in the comprehensive dataset according to daily time period;

[0021] determining peak hours according to the difference between weekdays and weekends;

[0022] analyzing activity type concentration and marking key focus areas;

[0023] clustering the behavior pattern in combination with weather conditions to generate the behavior pattern dataset.

[0024] Optionally, the prediction of demand from the behavior pattern dataset and the generation of the demand prediction dataset comprises:

[0025] integrating the behavior pattern data with historical usage records and the course activity schedule of the day into an initial dataset;

[0026] adopting a long short-term memory network to train a time series prediction model;

[0027] using the model to predict the number of participants in each area and time period in the coming week;

[0028] The demand prediction dataset is generated after bias correction of the prediction results.

[0029] Optionally, the calculating weather suitability score according to the demand prediction dataset and the weather data and generating the preliminary facility layout dataset comprises:

[0030] The demand prediction value is normalized and weighted with illumination, temperature, rainfall and wind direction data to calculate an initial weather score;

[0031] The score below the threshold is adjusted by linear interpolation;

[0032] A clustering algorithm is used to determine the region category and calculate the distance priority of the candidate location point to the cluster center;

[0033] The preliminary facility layout dataset is generated according to the priority and the adjusted score.

[0034] Optionally, the weather condition screening according to the preliminary facility layout dataset and generating the adjusted candidate location dataset comprises:

[0035] The location points meeting the preset threshold in illumination intensity, rainfall, temperature range and wind direction stability are sequentially screened;

[0036] The weather suitability score of the screened location points is recalculated;

[0037] The adjusted candidate location dataset is generated by sorting according to the score.

[0038] Optionally, the spatial analysis according to the adjusted candidate location dataset combined with the noise level and generating the optimized location dataset comprises:

[0039] The location point data is fused with the regional average decibel value to form an initial fusion dataset;

[0040] The noise level of each location point is calculated after division by region;

[0041] The location points with noise below the threshold are screened and the comprehensive score of weather and noise is calculated;

[0042] The optimized location dataset is generated after clustering optimization of the location points.

[0043] Optionally, the facility configuration according to the optimized location dataset combined with the facility cost and generating the facility layout planning dataset comprises:

[0044] The location data is spatially clustered to divide regions;

[0045] The required number of facilities in each region is initially allocated according to the facility coverage range;

[0046] adjusting the number of facilities using linear regression in combination with construction cost;

[0047] determining type combinations according to facility type suitability and maintenance cost;

[0048] generating the facility layout planning dataset containing the number of regional facilities and types.

[0049] In another aspect, the present application also provides an electronic device, comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0050] In another aspect, the present application also provides a computer readable storage medium, which stores a computer program, wherein the computer program is executable on a processor to implement the method.

[0051] Compared with the prior art, the present application has the following advantages and technical effects:

[0052] The application discloses a campus intelligent facility layout optimization method, which comprises the following steps: collecting and analyzing multi-source data, constructing a comprehensive dataset and a behavior pattern dataset, predicting future demand by using a long short-term memory network, evaluating the suitability of candidate positions in combination with meteorological conditions, and performing spatial analysis by considering noise levels, and finally generating an optimized facility layout plan. The method can dynamically adjust facility configuration and continuously optimize the matching degree of supply and demand, thereby effectively improving the utilization efficiency of campus resources. The application integrates multi-dimensional information such as course arrangement, activity records and meteorological data, realizes accurate prediction of campus flow and activity distribution and intelligent planning of facility layout, provides a data-driven scientific basis for campus management decision-making, and helps to create a more comfortable and efficient campus environment. BRIEF DESCRIPTION OF DRAWINGS

[0053] The accompanying drawings, which form a part of this application, are intended to provide further understanding of the application and are incorporated herein in their entirety. The schematic embodiments of the application and their descriptions are used to explain the application and do not constitute an improper limitation on the application. In the drawings:

[0054] Figure 1 The accompanying drawings, which form a part of this application, are intended to provide further understanding of the application and are incorporated herein in their entirety. The schematic embodiments of the application and their descriptions are used to explain the application and do not constitute an improper limitation on the application. In the drawings: DETAILED DESCRIPTION

[0055] It should be noted that the embodiments and features in the present application can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0056] It should be noted that the steps shown in the flowchart of the drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that here.

[0057] Embodiment one

[0058] As Figure 1 shown, the embodiment provides an outdoor space public facility layout optimization method based on big data analysis, including:

[0059] Obtain campus multi-source data and generate a comprehensive data set;

[0060] Extract behavior patterns from the comprehensive data set and generate a behavior pattern data set;

[0061] Predict demand from the behavior pattern data set and generate a demand prediction data set;

[0062] Calculate weather suitability scores from the demand prediction data set and weather data and generate a preliminary facility layout data set;

[0063] Perform weather condition screening from the preliminary facility layout data set and generate an adjusted candidate location data set;

[0064] Perform spatial analysis from the adjusted candidate location data set in combination with noise levels and generate an optimized location data set;

[0065] Calculate facility configuration from the optimized location data set in combination with facility costs and generate a facility layout planning data set;

[0066] Calculate supply-demand matching degree from the facility layout planning data set in combination with updated demand prediction data set and generate an updated configuration data set;

[0067] Recalculate facility configuration from the updated configuration data set in combination with historical maintenance records and spatial restrictions when supply-demand matching degree is below a threshold value and generate a final optimized facility layout data set.

[0068] Specifically includes:

[0069] Step S101, obtain course arrangement data, activity schedule records and weather condition data from the campus management system, including time stamp, location coordinates, number of participants, activity type, light intensity, wind direction, temperature and rainfall, construct a multi-source data collection framework, perform time stamp unification to international standardization organization format, numerical normalization and missing value filling, store to a unified database, generate a comprehensive data set containing time, location, activity type, number of participants and weather conditions.

[0070] Acquire course schedule, activity schedule and weather data, extract timestamp, location coordinate, number of participants, activity type, light intensity, wind direction, temperature and rainfall from campus management system and weather interface, store to temporary dataset. For the timestamp of temporary dataset, if the timestamp format is not international standard organization format, adopt time analysis function to convert to international standard organization format, obtain unified timestamp dataset. According to the unified timestamp dataset, adopt minimum-maximum normalization method, execute numerical normalization on the number of participants, light intensity, temperature and rainfall, obtain normalized dataset. For the normalized dataset, if there is missing value, adopt k-neighborhood algorithm to fill in the missing value, deduce the missing value based on the adjacent data of location coordinate and timestamp, obtain complete dataset. Extract timestamp, location coordinate, activity type, number of participants, light intensity, wind direction, temperature and rainfall from the complete dataset, store to unified database through database write interface, obtain storage dataset. According to the storage dataset, adopt clustering analysis method, group according to timestamp and location coordinate, merge activity type, number of participants and weather condition, generate comprehensive dataset. For the comprehensive dataset, adopt time series analysis, detect the change trend of timestamp and number of participants, weather condition, obtain activity and environment correlation mode.

[0071] Specifically, the course schedule data, activity schedule records, and weather condition data are obtained from the campus management system, a multi-source data collection framework is designed, and the course schedule is extracted from the campus management system through the API interface, including the timestamp 2025-07-16T08:00:00Z, location coordinates (31.2304, 121.4737), number of participants 50, and activity type “lecture”. At the same time, records are obtained from the activity schedule database, such as 2025-07-16T14:00:00Z, location (31.2310, 121.4740), number of people 100, and type “sports meeting”. Weather data is obtained through the weather station API, including time 2025-07-16T08:00:00Z, light intensity 500 lux, wind direction northeast (45°), temperature 25.5°C, and rainfall 0 mm. Data collection uses Python scripts, sends HTTP requests through the requests library, parses JSON responses, and stores them as Pandas DataFrames. The timestamp is converted to ISO8601 format, parsed using datetime.strptime, and output as YYYY-MM-DDThh:mm:ssZ. Numerical normalization uses the Min-Max algorithm, for example, the number of participants ranges from 0 to 200, and 50 is normalized to (50-0) / (200-0) = 0.25, and the temperature range is [0, 40], and 25.5 is normalized to 0.6375. Missing values are filled using the KNN algorithm (k = 3), and the nearest neighbor is calculated using the Euclidean distance, and the missing temperature is filled with the average of the neighboring records, such as (24.0, 26.0, 25.0) average 25.0. Data is stored in a MySQL database, the table structure includes fields time, location_x, location_y, event_type, participants, light, wind_direction, temperature, rainfall, and data is inserted using SQLAlchemy. A comprehensive data set is generated, multiple source DataFrames are merged, sorted by timestamp, and output example records: 2025-07-16T08:00:00Z, (31.2304, 121.4737), lecture, 0.25, 500 lux, 45°, 0.6375, 0 mm.

[0072] In step S102, based on the comprehensive data set, the number of participants, activity type distribution, and weather condition changes in each time period are extracted, clustering analysis and time series analysis are applied, and a behavior pattern data set is generated, including the peak period and main activity type of each region on weekdays and weekends.

[0073] The number of participants and activity type data of each region are obtained from the comprehensive data set, and a time period processing method is used to divide the distribution characteristics of different time periods in a day, and a preliminary time period division result is obtained. According to the time period division result, the number of participants in each region on weekdays and weekends is obtained, and the high and low fluctuation trend of the number of participants in each period is determined in combination with the distribution characteristics of the peak period. According to the high and low fluctuation trend of the number of participants, the distribution data of the activity type is obtained, the correlation between the activity type and the peak period is analyzed, and the concentration of the activity type in the distribution of different regions is judged. If the concentration of the activity type in the distribution of a region is higher than a preset threshold, the region is marked as a key attention region, its meteorological condition data is obtained, and the influence weight of the meteorological condition on the behavior mode is determined. Through the influence weight of the behavior mode, combined with the change data of the meteorological condition, a clustering analysis method is used to classify the behavior modes of each region, and a classified behavior mode set is obtained. According to the classified behavior mode set, the peak period data of weekdays and weekends is obtained, the cross characteristics of regional distribution and time period division are analyzed, and the final behavior mode distribution rule is determined.

[0074] Specifically, based on the analysis of the comprehensive data set, we first extract the number of participants, activity type distribution and meteorological condition change data of each region in the database for each time period. Assuming that the number of participants in a certain city A region is 500 in the morning of 8 o'clock on weekdays, the activity type is mainly commuting, accounting for 70%, running for fitness accounts for 20%, and other activities accounts for 10%, and the meteorological condition is sunny with a temperature of 20℃. While in the same time period on weekends, the number of participants drops to 300, the activity type is mainly leisure walking, accounting for 60%, running for fitness accounts for 30%, and other activities accounts for 10%, and the meteorological condition is cloudy with a temperature of 18℃. Then, we apply the K-means clustering algorithm to process the data, taking the number of participants, activity type proportion and meteorological condition as feature vectors, setting K value to 3, and iteratively calculating the Euclidean distance of each region data point to the cluster center. It is concluded that the A region belongs to the "commuting peak" cluster in the morning of 8 o'clock on weekdays, and belongs to the "leisure low peak" cluster on weekends. The clustering result shows that the peak period on weekdays is concentrated in 7-9 and 17-19, and the main activity is commuting, while the peak period on weekends is 10-12, and the main activity is leisure walking. Subsequently, an ARIMA time series analysis model is used to predict the number of participants on each day, with parameters p=1, d=1, q=1. The fluctuation trend of the number of participants in the A region at 8 o'clock in the future week is analyzed, and the prediction result shows that the number of participants is stable between 480-520 on weekdays, while the number of participants on weekends fluctuates greatly, with a prediction range of 280-320. Combined with the change of meteorological conditions such as temperature drop of 2℃, it is inferred that the number of participants will decrease by 5%. Finally, the behavior pattern data set is generated, which integrates the clustering and time series analysis results, records the peak period of A region on weekdays as 7-9 and 17-19, the main activity type as commuting accounting for 70%, the peak period on weekends as 10-12, the main activity type as leisure walking accounting for 60%, and associates the business rules such as the number of participants decreasing by 10% when the temperature is lower than 15℃, forming a complete behavior pattern description.

[0075] Step S103, based on the behavior pattern data set, combining the historical usage records of the past year and the course and activity arrangement of the day, a long short-term memory network model is constructed to predict the number of participants in each region at different time periods in the next week, and a demand prediction data set is generated, including region, time period and predicted number of participants.

[0076] The historical behavior pattern data, the past one-year use records, and the day's course and activity arrangement are acquired to construct an initial data set containing areas, time periods, and the number of participants. A long short-term memory network is used to train the initial data set to generate a time series prediction model to capture the trend of the number of participants in the areas and time periods. Through the time series prediction model, the number of participants in each area at different time periods in the future week is predicted in combination with the day's course and activity arrangement to obtain a preliminary prediction result. If the preliminary prediction result deviates from the historical behavior pattern by more than a preset threshold, the model parameters are adjusted, the time series prediction model is retrained, and an optimized prediction result is obtained. According to the optimized prediction result, a demand prediction data set containing areas, time periods, and the predicted number of participants is generated. For the demand prediction data set, grouping is performed according to the areas and time periods, the average number of participants of each group is calculated, and grouping statistical results are obtained. The distribution rule of the number of participants in the areas and time periods is extracted from the grouping statistical results to generate a final demand prediction data set.

[0077] Specifically, based on the behavior pattern dataset, combined with the historical usage records of the past year and the course and activity arrangement of the day, a long short-term memory network model is constructed to predict the number of participants in each area at different time periods in the next week, and a demand prediction dataset is generated. First, collect historical data for the past year, assuming there are 10 areas, and record the number of participants per hour per day for each area, forming a data matrix containing 365 days x 24 hours x 10 areas, totaling 87600 records, with data format as area ID, date, hour, and number of participants, for example, 50 people participated in area 1 at 8 o'clock on January 1, 2023. Next, integrate the course and activity arrangement data for the day, assuming there are 5 courses and 3 activities, each with specific time periods and estimated number of participants, such as course A in area 2 from 9 to 11 o'clock with an estimated 30 participants. By merging these data with historical data, a feature set is formed, including time, area, historical number of participants, activity type, and estimated number of participants. Then, a long short-term memory network model is constructed, with a 3-layer LSTM structure, each layer having 64 hidden units, input features being 24-hour data sequences for the past 7 days, i.e. 168 time steps, and output being the predicted number of participants per hour for the next 7 days, using mean squared error as the loss function, training the model using the Adam optimizer, with the training set being the first 80% of the data, i.e. 70080 records, and the validation set being the last 20%, i.e. 17520 records, training the model for 50 rounds until it converges, with a validation set error of 5.3%. Subsequently, the trained model is used to predict data for the next week, inputting the latest 7-day historical data and the course arrangement for the next week, and outputting the predicted number of participants per hour for each area, for example, 45 people are predicted to participate in area 1 at 8 o'clock on the first day of the future. Finally, the demand prediction dataset is generated, containing fields such as area ID, date, hour, and predicted number of participants, totaling 10 areas x 7 days x 24 hours = 1680 predicted records, saved as a CSV format file for subsequent analysis and resource allocation.

[0078] Step S104, based on the predicted number of participants in the demand prediction dataset, combined with the light intensity, wind direction, temperature, and rainfall in the comprehensive dataset, calculate the weather suitability score of each area, generate the preliminary facility layout dataset, containing candidate location points and their weather suitability scores.

[0079] The number of participants prediction value is obtained from the demand prediction dataset, the prediction value is normalized by using a standardization processing method, and the normalized number of participants prediction value is obtained. According to the normalized number of participants prediction value, the light intensity, wind direction, temperature and rainfall are obtained from the comprehensive dataset, the initial meteorological score of each region is calculated by using a weighted average method, and the initial meteorological score of each region is obtained. If the initial meteorological score is lower than a preset threshold value, the light intensity, wind direction, temperature and rainfall are adjusted by using a linear interpolation method, the adjusted meteorological score of each region is recalculated, and the adjusted meteorological score is obtained. According to the adjusted meteorological score, the K-means clustering algorithm is used to group each region, the clustering category of each region is determined, and the region clustering result is obtained. According to the region clustering result, the candidate location points are extracted from the comprehensive dataset, the distance weighted method is used to calculate the distance between each candidate location point and the cluster center, and the priority ranking of the candidate location points is obtained. Through the priority ranking of the candidate location points, combined with the adjusted meteorological score, a dataset containing the candidate location points and the meteorological suitability score is generated, and the preliminary facility layout dataset is obtained. If the number of candidate location points in the preliminary facility layout dataset is less than a preset threshold value, new candidate location points are supplemented from the comprehensive dataset, and the distance weighted method is repeated to calculate, and the updated preliminary facility layout dataset is obtained.

[0080] Specifically, based on the predicted number of participants in the demand prediction dataset, combined with the light intensity, wind direction, temperature and rainfall in the comprehensive dataset, the meteorological suitability score of each region is calculated, and the preliminary facility layout dataset is generated, including candidate location points and their scores. First, assuming that the demand prediction dataset contains the predicted number of participants in activity areas A, B and C as 500, 300 and 200 respectively, the number of people is converted into weight through normalization processing, and the calculation method is: regional weight = regional number of people / total number of people, and the weights of A, B and C are 0.5, 0.3 and 0.2 respectively. Next, extract the meteorological data from the comprehensive dataset, and assume that the light intensity of area A is 800 lux, the wind direction is north wind (0°), the temperature is 25℃, and the rainfall is 0 mm; the light intensity of area B is 600 lux, the wind direction is east wind (90°), the temperature is 22℃, and the rainfall is 2 mm; the light intensity of area C is 400 lux, the wind direction is south wind (180°), the temperature is 20℃, and the rainfall is 5 mm. The meteorological suitability score adopts a weighted comprehensive scoring algorithm, and the scoring formula is: suitability score = w1 x light score + w2 x wind score + w3 x temperature score + w4 x rainfall score, where the weights w1, w2, w3 and w4 are 0.3, 0.2, 0.3 and 0.2 respectively. The light score is calculated by linear normalization, and the formula is: (light intensity - minimum light) / (maximum light - minimum light), taking 800 and 400 lux as the upper and lower limits, and the light scores of A, B and C are 1, 0.5 and 0 respectively. The wind direction score is based on the definition of activity suitability, and the north wind is 1, the east wind is 0.5, and the south wind is 0, so the wind direction scores of A, B and C are 1, 0.5 and 0 respectively. The temperature score takes 20-25℃ as the optimum, and adopts the Gaussian function: exp(-((T-22.5)^2) / (2×2^2)), and the temperature scores of A, B and C are 0.88, 1 and 0.88 respectively. The rainfall score takes 0 mm as the optimum, and adopts the inverse function: 1 / (1+rainfall), and the rainfall scores of A, B and C are 1, 0.33 and 0.17 respectively. After comprehensive calculation, the suitability score of area A is 0.964, the suitability score of area B is 0.616, and the suitability score of area C is 0.298. Combined with the weight of the predicted number of participants, the final score is adjusted: final score = suitability score x number of participants weight, and the final scores of A, B and C are 0.482, 0.185 and 0.06 respectively. The preliminary facility layout dataset is generated, including candidate location points (A, B and C) and their scores (0.482, 0.185 and 0.06), and stored in JSON format: [{“location“:“A",“score“:0.482},{“location“:“B",“score“:0.185},{“location“:“C",“score“:0.06}].

[0081] In step S105, based on the preliminary facility layout dataset, candidate location points with illumination intensity lower than 1000 watts per square meter, rainfall lower than 5 millimeters per hour, temperature within 15 to 30 degrees Celsius, and stable wind direction are screened to generate an adjusted candidate location dataset, including the screened location points and their weather suitability scores.

[0082] The illumination intensity, rainfall, temperature range, and wind direction stability data are obtained from the facility layout dataset to generate an initial dataset. If the illumination intensity is lower than a preset threshold, the corresponding location point data is retained to obtain a first screened dataset. If the rainfall in the first screened dataset is lower than a preset threshold, the corresponding location point data is retained to obtain a second screened dataset. If the temperature in the second screened dataset is within a preset range, the corresponding location point data is retained to generate a third screened dataset. The K-means clustering algorithm is used to classify the wind direction stability in the third screened dataset to determine whether the wind direction stability meets a preset condition to obtain a fourth screened dataset. Based on the location points in the fourth screened dataset, the weather suitability scores are calculated to generate an adjusted candidate location dataset including the location points and scores. The scores in the adjusted candidate location dataset are sorted by a weighted summation method to obtain a final candidate location dataset.

[0083] Specifically, based on the screening and analysis of the preliminary facility layout dataset, we use information technology means to filter the candidate location points in multiple dimensions of meteorological conditions and generate an adjusted dataset. First, assume that the initial dataset contains 100 location points, each with attributes such as light intensity (unit: watts per square meter), rainfall (unit: millimeters per hour), temperature (unit: degrees Celsius), wind direction stability (represented by standard deviation, the smaller the value the more stable), etc. Taking light intensity as an example, set the threshold below 1000 watts per square meter. The system traverses the dataset and filters out location points that meet the conditions. Assume that among the initial data, 60 points have light intensity between 800 and 950 watts per square meter, meeting the requirements. Next, for the condition of rainfall below 5 millimeters per hour, the system further filters the above 60 points. Assume that among them, 50 points have rainfall between 0.5 and 4.8 millimeters per hour, meeting the conditions. Then, for the temperature range of 15 to 30 degrees Celsius, the system continues to filter the 50 points. Assume that finally, 40 points have a temperature between 18.5 and 28.3 degrees Celsius, meeting the standard. For wind direction stability, set the standard deviation threshold to 0.2. The system calculates the wind direction data fluctuation of each point. Assume that among the 40 points, 35 points have a standard deviation between 0.05 and 0.18, meeting the stability requirements. Finally, the system calculates the meteorological suitability score for the 35 filtered location points using a weighted scoring algorithm: light intensity accounts for 30% (standardized score = actual value / 1000*30), rainfall accounts for 20% (score = (5-actual value) / 5*20), temperature accounts for 30% (score = 1-|(actual value-22.5) / 7.5|*30), and wind direction stability accounts for 20% (score = (0.2-standard deviation) / 0.2*20). Assume that a certain location point has light intensity of 850, rainfall of 2, temperature of 25, and wind direction standard deviation of 0.1. Its total score is 25.5+12+27+10=74.5 points. After calculating the scores of all location points, the system generates an adjusted candidate location dataset containing location numbers and scores, forming a complete meteorological suitability assessment result.

[0084] Step S106, based on the adjusted candidate location dataset, combined with the average decibel value of each region, perform spatial analysis, filter the location points with average decibel value below 60 decibels, generate an optimized location dataset containing location points, meteorological suitability score and noise level.

[0085] The candidate location dataset and the regional average decibel value are acquired, the location point information, the meteorological suitability score and the noise level are integrated through a data fusion technology, and an initial fusion dataset is generated. If the location point information in the initial fusion dataset is complete and has no missing value, a spatial analysis method is used to divide the dataset by region to obtain a regionalized location dataset. According to the regionalized location dataset, the average decibel value of the location points in each region is calculated, and a regional noise distribution data is generated. If the average decibel value in the regional noise distribution data is lower than a preset threshold value, the corresponding location point information is retained to generate a preliminary screening dataset. Through the location point information in the preliminary screening dataset, the meteorological suitability score is fused, and a comprehensive score is calculated by using a weighted average method to obtain a comprehensive evaluation dataset. For the comprehensive evaluation dataset, a K-means clustering algorithm is used to classify the location points, and a classification optimization dataset is generated based on the meteorological suitability and the noise level. The location points meeting the screening conditions are extracted from the classification optimization dataset, the location point information, the meteorological suitability score and the noise level are integrated, and a final optimization dataset is generated.

[0086] Specifically, based on the adjusted candidate location dataset, it is assumed that there are 100 location points, each location point has latitude and longitude coordinates, a meteorological suitability score (0-100) and an initial noise level (decibel). First, missing values are processed through data cleaning, for example, a certain location point lacks noise data, then the average decibel value of the adjacent 10 location points is used for interpolation, such as location A lacking value, the average decibel value of the adjacent points is 65, and the noise level of A is supplemented to 65. Then, spatial analysis is performed in combination with the average decibel value of each region, a Kriging spatial interpolation algorithm is used, a search radius of 500 meters is set, and the weight is based on the reciprocal square of the distance, and the predicted decibel value of each location point is calculated.

[0087] For example, the decibel values of the adjacent points of location B are [62, 58, 60], the distances are [200, 300, 400] meters, and the predicted decibel value is calculated as (62 / 200 2 +58 / 300 2 +60 / 400 2 ) / (1 / 200 2 +1 / 300 2 +1 / 400 2 )=59.8. The points with an average decibel value lower than 60 are screened, the data set is traversed, and the points with a decibel value ≥60 are removed, it is assumed that there are 50 points remaining after screening, an optimization dataset is generated, which includes location point coordinates, meteorological suitability score and noise level.

[0088] For example, location C (longitude 120.1, latitude 30.2) has a decibel value of 55 and a meteorological score of 85, and is retained in the optimized dataset. To ensure the reliability of the results, cross-validation is used, with 80% of the data used to train the Kriging model and 20% used for verification. The root mean square error (RMSE) is calculated, and if RMSE < 5, the model is valid. The final dataset is output in JSON format, containing the fields: location_id, longitude, latitude, meteorological_score, noise_level, for example: {“location_id":“C",“longitude":120.1,“latitude":30.2,“meteorological_score":85,“noise_level":55}.

[0089] Step S107, based on the optimized location dataset, combined with facility coverage range, construction and maintenance cost, calculate the number and type of facilities in each region, generate facility layout planning dataset, including facility configuration list in each region.

[0090] By spatial clustering analysis of location data, using a pre-established classification model, the distribution characteristics of different regions are divided, and the preliminary regional division result is obtained. According to the regional division result, combined with the coverage range data, the coverage demand of facilities in each region is calculated, and the initial allocation scheme of the number of facilities in each region is determined. For the initial allocation scheme, the construction cost data is integrated, and the cost distribution is predicted using a linear regression model to determine whether the cost exceeds the preset threshold. If it exceeds, adjust the number of facilities to obtain an optimized allocation scheme. From the optimized allocation scheme, extract the number of facilities and regional division information, combined with facility type data, analyze the applicability of different types of facilities, and determine the combination configuration of facility types in each region. After obtaining the combination configuration of facility types, combined with the maintenance cost data, calculate the cost distribution of long-term operation, if the maintenance cost is higher than the preset threshold, then adjust the combination of facility types, get the final configuration list. According to the final configuration list, generate layout planning data containing regional facility distribution and type combination, complete the overall planning output of regional facilities.

[0091] Specifically, based on the optimized location dataset, assuming the dataset contains the longitude and latitude of 100 regions, population density (1000 to 5000 people per square kilometer), and traffic flow (1000 to 10000 vehicles per day). First, the regions are classified by K-means clustering algorithm (K=5), with the initial cluster centers randomly selected as the coordinates of 5 regions. The Euclidean distance of each region to the center is iteratively calculated (formula: √[(x2-x1)2+(y2-y1) 2]) until the clustering centers converge, obtaining 5 categories of regions.

[0092] For example, region A is high-density and high-traffic, which is classified as category A (about 20). Then, the facility coverage range is combined.

[0093] For example, the coverage radius of small facilities is 500 meters, medium facilities is 1000 meters, and large facilities is 2000 meters. The demand point coverage rate of each region is calculated, and a greedy algorithm is used to preferentially select large facilities that cover the most unserved demand points. The formula is: coverage rate = number of covered demand points / total number of demand points. Assuming that region A1 has 1000 demand points and large facilities cover 800 demand points, the coverage rate is 80%. Then, the construction and maintenance costs are considered (small facilities cost 500,000 yuan to build and 10,000 yuan per year to maintain, medium facilities cost 1,000,000 yuan to build and 20,000 yuan per year to maintain, and large facilities cost 2,000,000 yuan to build and 40,000 yuan per year to maintain). Through linear programming optimization, the total cost is minimized, and the objective function is: minimize Σ (construction cost i + maintenance cost i x years). The constraint condition is coverage rate ≥ 95%, and the years are set to 5. The solution is that region A1 is configured with 1 large facility and 2 small facilities, with a total cost of 200 + 2 × 50 + 5 × (40 + 2 × 10) = 450 million yuan. The facility layout planning dataset is generated, including region ID, facility type (small, medium, large), number, coordinates, covered demand point number, and total cost fields, such as {A1, large, 1, (x1, y1), 800, 450 million yuan}. Through analysis, A-class regions are preferentially configured with large facilities due to high density and high traffic, and other regions are configured with small and medium facilities according to demand and cost balance.

[0094] Step S108, based on the facility layout planning dataset, combined with the demand prediction dataset updated based on the latest course, activity, and weather data, the supply-demand matching degree of each region is calculated, the number and type of facilities are adjusted, and the updated configuration dataset is generated, including the adjusted facility configuration list.

[0095] By integrating facility layout and planning data, a basic dataset of regional distribution is constructed, the initial facility configuration information of each region is obtained, and the boundary range of regional division is determined. According to the course information, activity information and meteorological information, combined with the basic dataset, the random forest model in machine learning is used to predict the demand change trend of each region in different time periods, and the demand prediction result is obtained. According to the demand prediction result, the supply-demand matching degree of each region is calculated by fusing the regional division data, if the matching degree is lower than the preset threshold, the region is marked as a to-be-optimized region, and the list of regions to be adjusted is determined. Obtain the to-be-optimized region list, analyze the existing distribution of the number and type of facilities, if the existing number is insufficient to meet the predicted demand, increase the corresponding number of facilities, if the type does not match the predicted demand, adjust the facility type, and determine the adjusted configuration scheme. Through the adjusted configuration scheme, combined with the regional division data, the facility configuration details of each region are generated, and the preliminary optimized configuration list dataset is obtained. Using a data verification tool, the preliminary optimized configuration list dataset is detected for whether there is a configuration conflict or data missing, if an exception is detected, the adjustment scheme step is backtracked for correction, and the final updated configuration list is obtained.

[0096] Specifically, the supply-demand matching degree calculation based on the facility layout planning dataset and the demand prediction dataset and the facility configuration adjustment can be realized by the following specific method. First, assume that the facility layout planning dataset contains the existing facility distribution data of 10 regions in a city, each region has facilities such as sports fields, libraries and community centers, and the specific number is, for example, region A has 2 sports fields, 1 library and 3 community centers. Combined with the latest course, activity and weather data updated demand prediction dataset, the demand value of each region is obtained, for example, region A predicts demand for 3 sports fields, 2 libraries and 4 community centers, the data is predicted by a machine learning model, and the model is based on historical usage rate and weather influencing factors (such as reducing outdoor facility demand by 20% on rainy days) to calculate. Then, the supply-demand matching degree is calculated, using the formula: matching degree = 1-|existing number-predicted demand| / predicted demand, assuming that the sports field matching degree of region A is 1-|2-3| / 3=0.67, the library matching degree is 1-|1-2| / 2=0.5, and the community center matching degree is 1-|3-4| / 4=0.75, the average comprehensive matching degree is 0.64, if it is lower than the threshold 0.7, it needs to be adjusted. Then, adjust the number and type of facilities, based on the principle of preferentially increasing or decreasing facilities with low matching degree, while considering the regional population density and budget constraints (assuming the total budget is 5 million yuan, the construction cost of a sports field is 100,000 yuan, the construction cost of a library is 50,000 yuan, and the construction cost of a community center is 30,000 yuan), region A needs to increase 1 sports field, 1 library and 1 community center, with a total cost of 180,000 yuan within the budget. Finally, an updated configuration dataset is generated.

[0097] Step S109, based on the updated configuration dataset, if the supply-demand matching degree of any area is less than 80%, extract the historical maintenance records and regional space restrictions from the comprehensive dataset, recalculate the number and type of facilities, generate the final optimized facility layout dataset, including the facility configuration list of each area and the supply-demand matching degree.

[0098] From the updated configuration dataset, extract the supply-demand matching degree data of each area, and perform preliminary analysis for each area. If the supply-demand matching degree of a certain area is less than the preset threshold, mark the area as a to-be-optimized area, and obtain a to-be-optimized area list. For the to-be-optimized area list, obtain the corresponding historical maintenance records and regional space restriction information from the comprehensive dataset, and through data integration processing, associate the historical maintenance records with the space restriction information to determine the constraint condition set of each area. According to the constraint condition set, use a linear regression model to predict and calculate the number and type of facilities in each area to obtain a preliminary facility configuration scheme. Through space restriction verification of the preliminary facility configuration scheme, if the facility configuration of a certain area exceeds the space restriction range, adjust the number and type of facilities to obtain an adjusted facility configuration scheme. According to the adjusted facility configuration scheme, calculate the supply-demand matching degree value of each area. If the matching degree value is still lower than the preset threshold, return to the constraint condition set for iterative adjustment to obtain a final facility configuration list. For the final facility configuration list, generate an optimized layout dataset by integrating the facility configuration list of each area with the corresponding supply-demand matching degree value to determine the complete optimized layout result. Through a data storage tool, save the optimized layout result to a specified database for subsequent query and analysis, and complete the entire optimization process.

[0099] Specifically, based on the updated configuration dataset, the system will first automatically detect whether the supply-demand matching degree of each area is less than 80%.

[0100] For example, assuming the supply-demand match rate for region A is 75%, which is below a threshold, the system will trigger an optimization process. Next, the system extracts historical maintenance records and regional space constraint data for the region from the comprehensive dataset. Assume the historical records show that the facility failure rate in region A was 10% over the past year, and the space constraint is a maximum of 5 large facilities and 10 small facilities. The system then enters a phase of recalculating the number and type of facilities, using a linear programming algorithm. The objective function is set to maximize the supply-demand match rate, with constraints including space constraints and maintenance costs. Assuming each large facility covers a demand of 100 units, and each small facility covers 50 units, the total regional demand is 800 units. The optimal solution is calculated to be 4 large facilities and 8 small facilities, with a total coverage of 800 units, while keeping maintenance costs within budget. Following this, the system generates the final optimized facility layout dataset, containing a facility configuration list for region A (4 large facilities and 8 small facilities) and the updated supply-demand match rate. By calculating the ratio of coverage demand to total demand, the new match rate is determined to be 100%. To ensure logical rigor, the system also links business data, such as regional population density (assuming 500 people per square kilometer), to verify whether the facility configuration meets the actual demand growth trend, and the analysis shows that the configuration is reasonable.

[0101] Example 2

[0102] like Figure 1 As shown, this embodiment provides a method for optimizing the layout of public facilities in outdoor spaces based on big data analysis, including:

[0103] The system obtains course schedule data, activity schedule records, and meteorological condition data from the campus management system, including timestamps, location coordinates, number of participants, activity type, light intensity, wind direction, temperature, and rainfall. It constructs a multi-source data acquisition framework, unifies the timestamps to the International Organization for Standardization (ISO) format, performs numerical normalization and missing value imputation, stores the data in a unified database, and generates a comprehensive dataset containing time, location, activity type, number of participants, and meteorological conditions.

[0104] Based on the comprehensive dataset, the number of participants, activity types, and weather conditions of each region are extracted for each time period of the day. Cluster analysis and time series analysis are applied to generate a behavioral pattern dataset, which includes the peak times and main activity types of each region on weekdays and weekends.

[0105] Based on a behavioral pattern dataset, combined with historical usage records from the past year and the daily course and activity schedule, a long short-term memory network model is constructed to predict the number of participants in different regions at different time periods in the coming week, generating a demand prediction dataset that includes region, time period, and predicted number of participants.

[0106] Based on the predicted number of participants in the demand prediction dataset, combined with the light intensity, wind direction, temperature and rainfall in the comprehensive dataset, the weather suitability score of each region is calculated, and the preliminary facility layout dataset is generated, including candidate location points and their weather suitability scores;

[0107] Based on the preliminary facility layout dataset, candidate location points with light intensity below 1000 watts per square meter, rainfall below 5 millimeters per hour, temperature between 15 and 30 degrees Celsius, and stable wind direction are screened, and the adjusted candidate location dataset is generated, including the screened location points and their weather suitability scores;

[0108] Based on the adjusted candidate location dataset, combined with the average decibel value of each region, spatial analysis is performed to screen location points with average decibel value below 60 decibels, and the optimized location dataset is generated, including location points, weather suitability scores and noise levels;

[0109] Based on the optimized location dataset, combined with facility coverage, construction and maintenance costs, the number and type of facilities in each region are calculated, and the facility layout planning dataset is generated, including the facility configuration list of each region;

[0110] Based on the facility layout planning dataset, combined with the demand prediction dataset updated based on the latest courses, activities and weather data, the supply-demand matching degree of each region is calculated, and the number and type of facilities are adjusted, and the updated configuration dataset is generated, including the adjusted facility configuration list;

[0111] Based on the updated configuration dataset, if the supply-demand matching degree of any region is less than 80%, the historical maintenance records and regional space restrictions are extracted from the comprehensive dataset, and the number and type of facilities are recalculated, and the final optimized facility layout dataset is generated, including the facility configuration list and supply-demand matching degree of each region.

[0112] Further, the course arrangement data, activity schedule records and weather condition data from the campus management system include timestamp, location coordinates, number of participants, activity type, light intensity, wind direction, temperature and rainfall, a multi-source data collection framework is constructed, timestamp is unified to international standardization organization format, numerical normalization and missing value filling are performed, and stored in a unified database, generating a comprehensive dataset including time, location, activity type, number of participants and weather conditions, including:

[0113] Obtain course arrangement, activity schedule and weather data, extract timestamp, location coordinates, number of participants, activity type, light intensity, wind direction, temperature and rainfall from the campus management system and weather interface, and store them in a temporary dataset;

[0114] For the timestamp of the temporary dataset, if the timestamp format is not the international standardization organization format, a time resolution function is used to convert the timestamp to the international standardization organization format to obtain a unified timestamp dataset;

[0115] According to the unified timestamp dataset, a minimum-maximum normalization method is used to perform numerical normalization on the number of participants, light intensity, temperature, and rainfall to obtain a normalized dataset;

[0116] For the normalized dataset, if there are missing values, a k-nearest neighbor algorithm is used to fill in the missing values based on the adjacent data of the location coordinates and the timestamp to obtain a complete dataset;

[0117] From the complete dataset, the timestamp, location coordinates, activity type, number of participants, light intensity, wind direction, temperature, and rainfall are extracted, and stored in a unified database through a database write interface to obtain a storage dataset;

[0118] According to the storage dataset, a clustering analysis method is used to group by timestamp and location coordinates, and merge activity type, number of participants, and weather conditions to generate a comprehensive dataset;

[0119] For the comprehensive dataset, time series analysis is used to detect the trend of the timestamp and the number of participants and weather conditions to obtain the activity and environment correlation mode.

[0120] Further, based on the comprehensive dataset, the number of participants, activity type distribution, and weather condition changes of each region in each daily time period are extracted, and clustering analysis and time series analysis are applied to generate a behavior pattern dataset, which contains the peak period and main activity type of each region on weekdays and weekends, including:

[0121] By obtaining the number of participants and activity type data of each region from the comprehensive dataset, a time period processing method is used to divide the distribution characteristics of each daily time period to obtain a preliminary time period division result;

[0122] According to the time period division result, the number of participants of each region on weekdays and weekends is obtained, and combined with the distribution characteristics of the peak period, the high and low fluctuation trend of the number of participants in each period is determined;

[0123] For the high and low fluctuation trend of the number of participants, the distribution data of the activity type is obtained, the correlation between the activity type and the peak period is analyzed, and the concentration of the activity type in the distribution of different regions is determined;

[0124] If the concentration of the activity type in the distribution of a certain region is higher than a preset threshold, the region is marked as a key attention region, its weather condition data is obtained, and the influence weight of the weather condition on the behavior pattern is determined;

[0125] By the influence weight of the behavior mode, combined with the change data of the weather condition, the behavior modes of each region are classified by using the clustering analysis method to obtain a classified behavior mode set;

[0126] According to the classified behavior mode set, peak period data of working days and weekend days are obtained, the cross characteristics of regional distribution and time period division are analyzed, and the final behavior mode distribution rule is determined.

[0127] Further, based on the behavior mode data set, combined with the historical usage records of the past year and the course and activity arrangement of the day, a long short-term memory network model is constructed to predict the number of participants in each region at different time periods in the next week, and a demand prediction data set is generated, including region, time period and predicted number of participants, including:

[0128] The historical behavior mode data, the usage records of the past year and the course and activity arrangement of the day are obtained, and an initial data set containing region, time period and number of participants is constructed;

[0129] The long short-term memory network is used to train the initial data set to generate a time series prediction model to capture the trend of the number of participants in the region and time period;

[0130] Through the time series prediction model, combined with the course and activity arrangement of the day, the number of participants in each region at different time periods in the next week is predicted to obtain a preliminary prediction result;

[0131] If the deviation between the preliminary prediction result and the historical behavior mode exceeds the preset threshold, the model parameters are adjusted, the time series prediction model is retrained, and an optimized prediction result is obtained;

[0132] According to the optimized prediction result, a demand prediction data set containing region, time period and predicted number of participants is generated;

[0133] According to the demand prediction data set, grouping is performed according to region and time period, the average number of participants of each group is calculated, and a grouping statistical result is obtained;

[0134] The distribution rule of the number of participants in the region and time period is extracted from the grouping statistical result, and a final demand prediction data set is generated.

[0135] Further, based on the predicted number of participants in the demand prediction data set, combined with the light intensity, wind direction, temperature and rainfall in the comprehensive data set, the weather suitability score of each region is calculated, and a preliminary facility layout data set is generated, containing candidate location points and their weather suitability scores, including:

[0136] The predicted value of the number of participants is obtained from the demand prediction data set, and the predicted value is normalized by using a standardization processing method to obtain a normalized predicted value of the number of participants;

[0137] According to the normalized number of participants prediction value, the light intensity, wind direction, temperature and rainfall are obtained from the comprehensive data set, and the initial meteorological score of each region is calculated by using the weighted average method, and the initial meteorological score of each region is obtained;

[0138] If the initial meteorological score is lower than the preset threshold value, the light intensity, wind direction, temperature and rainfall are adjusted by using the linear interpolation method, and the adjusted meteorological score of each region is recalculated, and the adjusted meteorological score is obtained;

[0139] According to the adjusted meteorological score, the K-means clustering algorithm is used to group each region, and the clustering category of each region is determined, and the region clustering result is obtained;

[0140] According to the region clustering result, the candidate position points are extracted from the comprehensive data set, and the distance weighted method is used to calculate the distance between each candidate position point and the cluster center, and the priority ranking of the candidate position points is obtained;

[0141] Through the priority ranking of the candidate position points, combined with the adjusted meteorological score, a data set containing candidate position points and meteorological suitability score is generated, and a preliminary facility layout data set is obtained;

[0142] If the number of candidate position points in the preliminary facility layout data set is less than the preset threshold value, new candidate position points are supplemented from the comprehensive data set, and the distance weighted method is repeated to calculate, and an updated preliminary facility layout data set is obtained.

[0143] Further, based on the preliminary facility layout data set, the candidate position points with light intensity lower than 1000 watts per square meter, rainfall lower than 5 millimeters per hour, temperature in 15-30 degrees Celsius and stable wind direction are screened, and an adjusted candidate position data set is generated, containing the screened position points and meteorological suitability score, including:

[0144] The light intensity, rainfall, temperature range and wind direction stability data are obtained from the facility layout data set to generate an initial data set;

[0145] If the light intensity is lower than the preset threshold value, the corresponding position point data is retained to obtain a first screening data set;

[0146] If the rainfall in the first screening data set is lower than the preset threshold value, the corresponding position point data is retained to obtain a second screening data set;

[0147] If the temperature in the second screening data set is within the preset range, the corresponding position point data is retained to generate a third screening data set;

[0148] The K-means clustering algorithm is used to classify the wind direction stability in the third screened data set, to determine whether the wind direction stability meets the preset condition, and to obtain a fourth screened data set;

[0149] According to the position points in the fourth screened data set, the weather suitability score is calculated, and an adjusted candidate position data set containing position points and scores is generated;

[0150] The scores in the adjusted candidate position data set are sorted by a weighted summation method, and a final candidate position data set is obtained.

[0151] Further, based on the adjusted candidate position data set, the average decibel value of each region is combined to perform spatial analysis, and position points with an average decibel value lower than 60 decibels are screened to generate an optimized position data set containing position points, weather suitability scores and noise levels, including:

[0152] The candidate position data set and the average decibel value of the region are obtained, and the position point information, weather suitability score and noise level are integrated by data fusion technology to generate an initial fusion data set;

[0153] If the position point information in the initial fusion data set is complete and has no missing values, the data set is divided by region using a spatial analysis method to obtain a regionalized position data set;

[0154] According to the regionalized position data set, the average decibel value of the position points in each region is calculated to generate a regional noise distribution data;

[0155] If the average decibel value in the regional noise distribution data is lower than the preset threshold, the corresponding position point information is retained to generate a preliminary screening data set;

[0156] By fusing the weather suitability score with the position point information in the preliminary screening data set, a comprehensive score is calculated using a weighted average method to obtain a comprehensive evaluation data set;

[0157] For the comprehensive evaluation data set, the K-means clustering algorithm is used to classify the position points, and a classification optimization data set is generated based on the weather suitability and noise level;

[0158] From the classification optimization data set, position points that meet the screening conditions are extracted, and the position point information, weather suitability score and noise level are integrated to generate a final optimization data set.

[0159] Further, based on the optimized position data set, the facility coverage range, construction and maintenance cost are combined to calculate the number and type of facilities in each region, and a facility layout planning data set containing a facility configuration list for each region is generated, including:

[0160] By spatial clustering analysis on the position data, the distribution characteristics of different regions are divided by using the pre-established classification model, and the preliminary regional division result is obtained.

[0161] According to the regional division result, combined with the coverage range data, the coverage demand of facilities in each region is calculated, and the initial allocation scheme of the number of facilities in each region is determined;

[0162] For the initial allocation scheme, the construction cost data is fused, the cost distribution is predicted by using the linear regression model, whether the cost exceeds the preset threshold is judged, if it exceeds, the number of facilities is adjusted, and the optimized allocation scheme is obtained;

[0163] From the optimized allocation scheme, the number of facilities and regional division information are extracted, combined with the facility type data, the applicability of different types of facilities is analyzed, and the combination configuration of facility types in each region is determined;

[0164] After obtaining the combination configuration of facility types, combined with the maintenance cost data, the cost distribution of long-term operation is calculated, if the maintenance cost is higher than the preset threshold, the facility type combination is re-adjusted, and the final configuration list is obtained;

[0165] According to the final configuration list, the layout planning data containing regional facility distribution and type combination is generated, and the comprehensive planning output of regional facilities is completed.

[0166] Further, based on the facility layout planning data set, combined with the demand prediction data set updated based on the latest course, activity and weather data, the supply-demand matching degree of each region is calculated, the number and type of facilities are adjusted, and the updated configuration data set is generated, including the adjusted facility configuration list, including:

[0167] By integrating the facility layout and planning data, the basic data set of regional distribution is constructed, the initial facility configuration information of each region is obtained, and the boundary range of regional division is determined;

[0168] According to the course information and activity information and weather information, combined with the basic data set, the random forest model in machine learning is used to predict the demand change trend of each region in different time periods, and the demand prediction result is obtained;

[0169] For the demand prediction result, combined with the regional division data, the supply-demand matching degree of each region is calculated, if the matching degree is lower than the preset threshold, the region is marked as a to-be-optimized region, and the region list to be adjusted is determined;

[0170] Obtain the to-be-optimized region list, analyze the existing distribution of the number of facilities and the type of facilities, if the existing number is insufficient to meet the predicted demand, increase the corresponding number of facilities, if the type does not match the predicted demand, adjust the facility type, and determine the adjusted configuration scheme;

[0171] By adjusting the configuration scheme, combined with the regional division data, the facility configuration details of each region are generated, and the preliminary optimized configuration list dataset is obtained;

[0172] Using a data verification tool, the preliminary optimized configuration list dataset is detected for configuration conflicts or data missing. If an exception is detected, the adjustment scheme step is traced back for correction, and the final updated configuration list is obtained.

[0173] Further, based on the updated configuration dataset, if the supply-demand matching degree of any region is less than 80%, the historical maintenance records and regional space restrictions are extracted from the comprehensive dataset to recalculate the number and type of facilities, and generate the final optimized facility layout dataset, including the facility configuration list and supply-demand matching degree of each region, including:

[0174] From the updated configuration dataset, the supply-demand matching degree data of each region is extracted, and a preliminary analysis is performed for each region. If the supply-demand matching degree of a region is less than a preset threshold, the region is marked as a to-be-optimized region, and a to-be-optimized region list is obtained;

[0175] For the to-be-optimized region list, the corresponding historical maintenance records and regional space restriction information are obtained from the comprehensive dataset. Through data integration processing, the historical maintenance records and space restriction information are associated to determine the constraint condition set of each region;

[0176] According to the constraint condition set, a linear regression model is used to predict and calculate the number and type of facilities in each region to obtain a preliminary facility configuration scheme;

[0177] Through space restriction verification of the preliminary facility configuration scheme, if the facility configuration of a region exceeds the space restriction range, the number and type of facilities are adjusted to obtain an adjusted facility configuration scheme;

[0178] According to the adjusted facility configuration scheme, the supply-demand matching degree value of each region is calculated. If the matching degree value is still lower than the preset threshold, return to the constraint condition set for iterative adjustment to obtain the final facility configuration list;

[0179] For the final facility configuration list, an optimized layout dataset is generated, which integrates the facility configuration list of each region and the corresponding supply-demand matching degree value to determine the complete optimized layout result;

[0180] Through a data storage tool, the optimized layout result is saved to a specified database for subsequent query and analysis, completing the entire optimization process.

[0181] In another aspect, the embodiment further provides an electronic device, comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, wherein the processor implements the method when executing the computing program.

[0182] In another aspect, the embodiment further provides a computer readable storage medium, which stores a computer program, wherein the computer program is executable on a processor to implement the method.

[0183] The above merely provides the preferred embodiments of the present application, but the protection scope of the present application is not limited thereto, and any changes or substitutions easily conceived by those skilled in the art within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for optimizing the layout of public facilities in outdoor spaces based on big data analysis, characterized in that, include: Acquire multi-source campus data and generate a comprehensive dataset; Extract behavioral patterns from the comprehensive dataset and generate a behavioral pattern dataset; Based on the behavioral pattern dataset, predict demand and generate a demand prediction dataset; Based on the demand forecast dataset and meteorological data, a meteorological suitability score is calculated and a preliminary facility layout dataset is generated; Based on the preliminary facility layout dataset, meteorological conditions are used to filter and an adjusted candidate location dataset is generated. Based on the adjusted candidate location dataset and the noise level, spatial analysis is performed to generate an optimized location dataset. Based on the optimized location dataset and the facility cost calculation, the facility configuration is generated and a facility layout planning dataset is produced. Based on the facility layout planning dataset and the updated demand forecast dataset, the supply and demand matching degree is calculated and an updated configuration dataset is generated. Based on the updated configuration dataset, when the supply-demand matching degree is below the threshold, the facility configuration is recalculated by combining historical maintenance records and space constraints, and the final optimized facility layout dataset is generated.

2. The method according to claim 1, characterized in that, The process of acquiring multi-source campus data and generating a comprehensive dataset includes: Collect course schedules, activity schedules, and weather data from the campus management system and weather interface; The timestamps of the data are standardized to ISO format and their values ​​are normalized. Missing values ​​are filled using the nearest neighbor algorithm and then stored in a unified database; The data in the database are clustered and merged according to time and location to generate the comprehensive dataset.

3. The method according to claim 1, characterized in that, The step of extracting behavioral patterns from the comprehensive dataset and generating a behavioral pattern dataset includes: The number of participants and activity types in the comprehensive dataset are divided according to daily time periods; Peak hours are determined based on the difference between weekdays and weekends; Analyze the concentration of activity types and mark key areas of focus; The behavior patterns are clustered based on meteorological conditions to generate the behavior pattern dataset.

4. The method according to claim 1, characterized in that, The step of predicting demand based on the behavioral pattern dataset and generating a demand prediction dataset includes: The behavioral pattern data was integrated with historical usage records and the daily course activity schedule into an initial dataset; The temporal prediction model is trained using a long short-term memory network; The model is used to predict the number of participants in different regions and time periods for the coming week; The demand forecast dataset is generated after bias correction of the forecast results.

5. The method according to claim 1, characterized in that, The process of calculating a meteorological suitability score and generating a preliminary facility layout dataset based on the demand forecast dataset and meteorological data includes: The demand forecast values ​​are normalized and then weighted with data on sunshine, temperature, rainfall, and wind direction to calculate the initial meteorological score; Linear interpolation is used to adjust scores below a threshold; Clustering algorithms are used to determine region categories and calculate the priority of distances between candidate locations and cluster centers; The preliminary facility layout dataset is generated based on the stated priorities and adjusted scores.

6. The method according to claim 1, characterized in that, The step of filtering meteorological conditions based on the preliminary facility layout dataset and generating an adjusted candidate location dataset includes: Locations that meet preset thresholds in terms of light intensity, rainfall, temperature range, and wind direction stability are selected sequentially. Recalculate the meteorological suitability score for the selected locations; The adjusted candidate location dataset is generated by sorting the scores.

7. The method according to claim 1, characterized in that, The step of performing spatial analysis based on the adjusted candidate location dataset and noise level to generate an optimized location dataset includes: The location point data and the regional average decibel value are merged into an initial fused dataset; Calculate the noise level at each location point after dividing the area; Select locations where noise levels are below a threshold and calculate a combined score for both weather and noise levels; The optimized location dataset is generated after clustering optimization of the location points.

8. The method according to claim 1, characterized in that, The step of calculating facility configuration and generating facility layout planning dataset based on optimized location dataset and facility cost includes: Spatial clustering of location data to divide regions; The initial allocation of the required number of facilities for each area is calculated based on the facility coverage area. The number of facilities was adjusted using linear regression, taking into account construction costs. The combination of types should be determined based on the applicability of facility types and maintenance costs; Generate a facility layout planning dataset containing the number and type of facilities in the area.

9. An electronic device comprising a memory, a processor, and a computing program stored in the memory and executable on the processor, characterized in that, When the processor executes the computing program, it implements the method of any one of claims 1-8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1-8.

Images