Intelligent information management method and system based on artificial intelligence

Abstract

The invention discloses an intelligent information management method and system based on artificial intelligence, and relates to the technical field of intelligent information management, and the method comprises the steps: obtaining a rainfall coefficient, and transmitting early warning information to a canteen management end when the rainfall coefficient is higher than a condition threshold value; predicting the dining hall transaction amount, and if the increment of the predicted dining hall transaction amount exceeds the expectation, sending out communication; predicting and generating a dish arrangement scheme of a current park canteen, and sending the dish arrangement scheme to a canteen management end; predicting the remaining dishes in the current prediction period, sending a supplement notification, and predicting the predicted remaining amount of each non-to-be-added dish at the end of the dining; and evaluating the non-to-be-added dishes with the possible residual amount, and according to the dish evaluation coefficient Cpx, performing corresponding discount strategies matched with the non-to-be-added dishes. The added dishes can be sold in a large probability, dish supply of the park is fully guaranteed, and the operation risk of the park is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent information management, and in particular to an intelligent information management method and system based on artificial intelligence. Background Art

[0002] With the further extension of urbanization and industrialization, there are more and more industrial parks in urban areas and their surrounding areas. In order to ensure the dining needs of park employees, park canteens are usually set up in the parks.

[0003] When it rains, the delivery of takeout will be greatly affected. If the takeout cannot be delivered in time, a large number of park employees will turn to the canteen for meals. At this time, the canteen may not prepare enough dishes, which may cause some park employees to have difficulty in eating in time at noon. However, if the canteen increases the amount of prepared meals than usual, the number of diners may be smaller, and some meals may not be sold in time, resulting in waste, and then creating operational risks.

[0004] With the development and application of artificial intelligence technology, the means of data analysis and processing are becoming more and more comprehensive. In order to solve the above problems, the present invention provides an information intelligent management method and system based on artificial intelligence. Summary of the invention

[0005] 1. Technical issues to be resolved In view of the deficiencies in the prior art, the present invention provides an information intelligent management method and system based on artificial intelligence, which predicts the canteen transaction amount. If the predicted increase in the canteen transaction amount exceeds expectations, a communication is sent out; the current dish arrangement plan of the park canteen is predicted and generated, and the dish arrangement plan is sent to the canteen management end; the remaining dishes in the current prediction period are predicted, and a supplementary notification is issued to predict the predicted remaining quantity of each non-to-be-added dish at the end of the meal; non-to-be-added dishes that may have remaining quantities are evaluated, and corresponding discount strategies are matched for the non-to-be-added dishes according to the dish evaluation coefficient Cpx, thereby solving the problems in the background technology.

[0006] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: An information intelligent management method based on artificial intelligence includes: when a campus cafeteria is in operation, if the external weather is in a rainy state, monitoring the weather conditions, establishing a weather condition data set based on the monitoring results and then obtaining a rainfall coefficient Jxs, and if the rainfall coefficient Jxs is higher than a condition threshold, sending a warning message to the cafeteria management end; including: Obtain the current weather data. If it is currently in a continuous rainfall state, obtain the current rainfall per hour and generate the average rainfall Pr; obtain the driving visibility Xd under the current conditions; when the average rainfall Pr exceeds the preset rainfall threshold, obtain the traffic congestion data in the rainfall area, identify the average driving speed of the vehicle, compare the speed with the speed under good weather conditions, and determine the ratio as the speed ratio Sd; summarize the average rainfall Pr, driving visibility Xd and speed ratio Sd to establish a weather condition data set; After receiving the warning information, the canteen dining model after training is used to predict the canteen transaction amount. If the predicted increase in the canteen transaction amount exceeds the expectation, a request message is sent out; if it does not exceed the expectation, the supply of some dishes is increased; After receiving the request information, the historical dining records of employees in the park cafeteria are obtained, and combined with the dining data of employees in each park, the trained dish allocation prediction model is used to predict and generate the current dish arrangement plan of the park cafeteria, and the dish arrangement plan is sent to the cafeteria management end; After obtaining and executing the dish arrangement plan, the trained dish ordering prediction model is used to predict the remaining dishes in the current prediction period. If the remaining proportion of dishes to be added is lower than expected, a supplementary notification is issued to predict the predicted remaining quantity of each non-dish to be added at the end of the meal; Evaluate the dishes that may have surplus quantity and are not to be added, generate the dish evaluation coefficient Cpx, match the corresponding discount strategy for the dishes that are not to be added according to the dish evaluation coefficient Cpx, and choose the response method according to the execution effect of the discount strategy, and handle the dishes that are not to be added and the dishes to be added separately on the next operating day.

[0007] Furthermore, the rainfall coefficient Jxs is generated from the weather condition data set, wherein the rainfall coefficient Jxs is generated as follows: after dimensionless processing of the average rainfall Pr, driving visibility Xd and speed ratio Sd, according to the following formula:

[0008] Weight coefficient: , ,and ; If the rainfall coefficient Jxs is higher than the preset condition threshold, an early warning message will be sent to the canteen management end.

[0009] Furthermore, after receiving the early warning information, the canteen management end obtains weather data in the past year, obtains dining data in the park canteen, and separates the daily total transaction amount Ze of the park canteen from the dining data; summarizes the dining records of the park canteen and the corresponding weather data, and uses a neural convolution model to establish and generate a canteen dining model after training and testing; under the current rainfall condition, uses the canteen dining model to predict the total transaction amount Ze to obtain the predicted transaction amount.

[0010] Furthermore, the total transaction amount Ze of the previous canteen operation day is obtained. If the predicted transaction amount is higher than the total transaction amount Ze, and the difference between the two is higher than the preset incremental threshold, a request message is sent to the outside; if it is not exceeded, the top several dishes with the highest sales volume are obtained from the dining records of the canteen, and they are determined as dishes to be added; after determining the dishes to be added, the sufficient supply of dishes in the park canteen can be ensured; after receiving the request message, the canteen management end increases the current supply of dishes to be added in the park canteen so that the increase in supply is consistent with the above-mentioned difference.

[0011] Furthermore, after receiving the request information, the dining data of all park employees in the most recent recording period is queried from the dining records of the park cafeteria to form a historical dining record of the employees; after receiving the warning information, the probability of each park employee going to the park cafeteria under the current weather conditions is obtained from the historical dining records to generate the dining probability; The dish preferences and dining probabilities of employees in each park are obtained from the employees' historical dining records, and a dining dataset is established after aggregation; a neural convolutional model is used to generate and output a dish allocation prediction model after training and testing using part of the data in the dining dataset; the current weather conditions, the dish preferences and dining probabilities of park employees are used as input, and the dish allocation prediction model is used to predict and analyze the dish arrangement of the park canteen on the day, and a dish arrangement plan for the park canteen is generated after iteration.

[0012] Furthermore, after receiving the menu arrangement plan and executing it, the canteen management end obtains the identity information of the park employees entering the park canteen and the current flow of people after the canteen enters the dining time period; uses the neural convolution model, combines the above information, generates a menu order prediction model after training and testing, and sets several continuous prediction cycles; During the current forecast period, the order prediction model is used to predict the remaining dishes in the campus cafeteria during the next forecast period and obtain the forecast results; if the remaining proportion of dishes to be added in the next forecast period is lower than the preset first proportion threshold, a dish replenishment notification is issued.

[0013] Furthermore, when the number of diners in the park canteen exceeds the preset threshold, the remaining proportion of each non-to-be-added dish in the park canteen within the current prediction period is obtained. If the remaining proportion is higher than the preset second proportion threshold, the current dining data in the park canteen is combined with the trained ordering prediction model to predict the remaining amount of non-to-be-added dishes in the canteen at the end of the meal in the park canteen, and generate the predicted remaining amount of each non-to-be-added dish.

[0014] Furthermore, the predicted remaining quantity of each dish that is not to be added is obtained, and each dish that is not to be added is evaluated to generate a dish evaluation coefficient Cpx; the dish evaluation coefficient Cpx is obtained, and the corresponding discount strategy is matched according to the distribution of the dish evaluation coefficient Cpx; the discount strategy is implemented for the dishes that are not to be added; after the current dining time is over, if there are still remaining dishes that are not to be added, the discount will be increased on the next operating day; if there are still remaining dishes to be added, the preparation quantity will be reduced.

[0015] Furthermore, the dish evaluation coefficient Cpx is obtained as follows: the dish price Cj and dish quantity Cv of each dish not to be added, as well as the sales volume of the dish on that day are obtained, and the dish sales volume Xw is generated from the sales volume; after dimensionless processing of the above parameters, according to the following formula:

[0016] The parameters have the following meanings: Price factor , is the food quantity factor, , is the sales factor, .

[0017] An information intelligent management system based on artificial intelligence, comprising: The monitoring unit monitors the weather conditions when the canteen is in operation and the external weather is raining. The monitoring unit establishes a weather condition data set based on the monitoring results and obtains the rainfall coefficient Jxs. If the rainfall coefficient Jxs is higher than the condition threshold, an early warning message is sent to the canteen management end. The analysis unit, after receiving the warning information, uses the trained canteen dining model to predict the canteen transaction amount. If the predicted increase in the canteen transaction amount exceeds the expectation, a request message is sent out; if it does not exceed the expectation, the supply of some dishes is increased; After receiving the request information, the plan formulation unit obtains the historical dining records of the employees in the park cafeteria, and combines the dining data of the employees in each park, uses the trained dish allocation prediction model, predicts and generates the current dish arrangement plan of the park cafeteria, and sends the dish arrangement plan to the cafeteria management end; The prediction unit, after obtaining and executing the dish arrangement plan, uses the trained order prediction model to predict the remaining dishes in the current prediction period. If the remaining proportion of dishes to be added is lower than expected, a supplementary notification is issued to predict the predicted remaining quantity of each non-dish to be added at the end of the meal; The evaluation unit evaluates the dishes that may have surplus quantity and are not to be added, generates the dish evaluation coefficient Cpx, matches the corresponding discount strategy for the dishes that are not to be added according to the dish evaluation coefficient Cpx, and selects the response method according to the execution effect of the discount strategy, and handles the dishes that are not to be added and the dishes to be added separately within the next operation day.

[0018] (III) Beneficial effects The present invention provides an information intelligent management method and system based on artificial intelligence, which has the following beneficial effects: 1. Judge the takeaway delivery conditions in the area where the park is located based on the rainfall coefficient Jxs. If the current rainfall gradually increases, the takeaway delivery environment will gradually deteriorate, and it is necessary to notify the park canteen in time. When the takeaway of a large number of park employees is difficult to deliver, more dishes need to be prepared than usual.

[0019] 2. Establish a canteen dining model and obtain predicted transaction volume to determine the dishes to be added. When the relevance is limited, give priority to adding several dishes with the best sales, so as to ensure that the added dishes can be sold with a high probability. While fully guaranteeing the supply of dishes in the park, it also reduces its own operational risks.

[0020] 3. Forecast the remaining quantity of dishes that are not to be added and dishes that are to be added. For dishes that are easy to sell and are to be added, when they are insufficient, they can be supplemented in appropriate amounts based on the insufficient quantity to fully guarantee the supply of dishes and improve the dining experience of employees in the park. For dishes that are difficult to sell and are not to be added, continuously predict their possible remaining quantity so that they can be processed in time to supplement the insufficient supply of dishes to be added.

[0021] 4. After generating the dish evaluation coefficient Cpx and giving the corresponding discount strategy, a marketing strategy is formed. When there may be surplus food in the campus canteen, it can be dealt with in time. At the same time, some dishes can be supplemented to reduce operational risks. After the end of the day's operation, feedback can be obtained to adjust the marketing strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a flow chart of the information intelligent management method based on artificial intelligence of the present invention; Figure 2 It is a schematic diagram of the structure of the information intelligent management system based on artificial intelligence of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] See also Figure 1 The present invention provides an information intelligent management method based on artificial intelligence, comprising: Step 1: When the canteen is in operation, if the external weather is raining, monitor the weather conditions, establish a weather condition data set based on the monitoring results, and then obtain the rainfall coefficient Jxs. If the rainfall coefficient Jxs is higher than the condition threshold, send an early warning message to the canteen management end; when the weather changes, it can be predicted in advance and handled in time; The step 1 includes the following contents: Step 101, obtaining current weather data, which can be obtained from a nearby weather station or through a weather forecast. If it is currently in a continuous rainfall state, the current rainfall per hour is obtained to generate an average rainfall Pr; the driving visibility Xd under the current conditions is obtained; When the average rainfall Pr exceeds a preset rainfall threshold, traffic congestion data in the rainfall area is obtained from a public channel, such as a navigation map, and an average driving speed of vehicles is identified, and the speed is compared with the speed under good weather conditions (i.e., clear weather conditions), and the ratio is determined as a speed ratio Sd; The average rainfall Pr, driving visibility Xd and speed ratio Sd are summarized to establish a weather condition data set; Step 102: Generate a rainfall coefficient Jxs from the weather condition data set, wherein the rainfall coefficient Jxs is generated as follows: after dimensionless processing of the average rainfall Pr, the driving visibility Xd and the speed ratio Sd, according to the following formula:

[0025] Weight coefficient: , ,and ; The method of obtaining the weight coefficient can refer to the hierarchical analysis method; Obtain the rainfall coefficient Jxs. If the rainfall coefficient Jxs is higher than the preset condition threshold, a warning message will be sent to the canteen management end. After the rainfall coefficient Jxs is generated, it can be judged whether the current weather conditions are conducive to food delivery based on the rainfall coefficient Jxs.

[0026] When using, combine the contents in steps 101 and 102: When it is detected that the weather conditions have changed and the rainy stage has begun, the average rainfall Pr is obtained and finally the rainfall coefficient Jxs is generated. The takeaway delivery conditions in the area where the park is located are judged based on the rainfall coefficient Jxs. If the current rainfall gradually increases, the takeaway delivery environment will gradually deteriorate. At this time, it is necessary to notify the park canteen in time. When the takeaway of a large number of park employees is difficult to deliver, more dishes need to be prepared than usual.

[0027] Step 2: After receiving the warning information, use the trained canteen dining model to predict the canteen transaction amount. If the predicted increase in the canteen transaction amount exceeds the expectation, a request message is sent out; if it does not exceed the expectation, the supply of some dishes is increased; The step 2 includes the following contents: Step 201: After receiving the warning information, the dining records and corresponding weather data of the campus cafeteria are summarized, and a cafeteria dining model is established and generated after training and testing using a neural convolution model; under the current rainfall condition, the cafeteria dining model is used to predict the total transaction amount Ze to obtain the predicted transaction amount; Step 202: Obtain the total transaction amount Ze of the previous canteen operation day. If the predicted transaction amount is higher than the total transaction amount Ze, and the difference between the two is higher than the preset increment threshold, send a request message to the outside; if it is not higher, obtain the first several dishes with the highest sales volume from the dining records of the canteen, such as the first three dishes, and determine them as dishes to be added; after determining the dishes to be added, it is possible to ensure the sufficient supply of dishes in the campus canteen; Step 203: After receiving the request information, the canteen management terminal increases the supply of the dishes to be added in the park canteen so that the supply increase is consistent with the above difference.

[0028] When using, combine the contents in steps 201 to 203: When in use, a canteen dining model is established and the predicted transaction amount is obtained. Based on the predicted results, several dishes with the best sales are increased in priority, thereby ensuring that the added dishes have a high probability of being sold. While fully guaranteeing the supply of dishes in the park, it also reduces its own operational risks.

[0029] Step 3: After receiving the request information, obtain the historical dining records of the employees in the park canteen, and combine the dining data of the employees in each park, use the trained dish allocation prediction model to predict and generate the current dish arrangement plan of the park canteen, and send the dish arrangement plan to the canteen management end; The step three includes the following contents: Step 301: after receiving the request information, query and obtain the dining data of all employees in the park in the most recent recording period, for example, a recording period can be one quarter or two quarters, to form a historical dining record of employees; after receiving the warning information, obtain the probability of each employee in the park going to the park canteen under the current weather conditions from the historical dining record to generate the dining probability; at this time, some people with the lowest dining probability can be excluded; Step 302: Obtain the food preferences and dining probabilities of employees in each park from the employee's historical dining records, and establish a dining data set after summarizing; use a neural convolution model, and generate and output a food distribution prediction model after training and testing using part of the data in the dining data set; Step 303: Using the current weather conditions, the food preferences and dining probabilities of the park employees as input, a food distribution prediction model is used to predict and analyze the food arrangement of the park canteen on the day, and a food arrangement plan for the park canteen is generated after iteration.

[0030] When using, combine the contents in steps 301 to 303: After receiving the request information, the dining data set of the park employees is used to train and generate a menu distribution prediction model. After determining the weather conditions, the probability of employees dining and their menu preferences, a menu arrangement plan is established on the premise of comprehensive data preparation. The park canteen can prepare meals on the same day based on the menu arrangement plan, thereby fully guaranteeing the preferences of employees in each park for dishes, and making the preparation quantity more accurate when preparing meals, further reducing its own operational risks while fully guaranteeing the meals of all park employees.

[0031] Step 4: After obtaining and executing the dish arrangement plan, use the trained dish ordering prediction model to predict the remaining dishes in the current prediction period. If the remaining proportion of dishes to be added is lower than expected, a supplementary notification is issued to predict the predicted remaining quantity of each non-dish to be added at the end of the meal; The step 4 includes the following contents: Step 401: After receiving the dish arrangement plan at the canteen management end, the plan is executed; after the canteen enters the dining time period, the identity information of the park employees entering the canteen for dining and the current flow of people information are obtained; Using the neural convolution model, combined with the above information, after training and testing, a dish ordering prediction model is generated, and several continuous prediction cycles are set along the time axis. For example, a prediction cycle is 10 minutes. When the meal preparation is initially completed, the dish ordering prediction model is trained and generated to predict the ordering process of the park employees. Step 402: In the current forecast period, the order prediction model is used to predict the remaining dishes in the campus cafeteria in the next forecast period, and the prediction result is obtained; if the remaining proportion of the dishes to be added in the next forecast period is lower than the preset first proportion threshold, a dish replenishment notification is issued; for the dishes to be added that are easy to sell, when the remaining quantity is insufficient, an appropriate amount can be replenished; Step 403, when the number of diners in the park canteen exceeds the preset number of diners threshold, obtain the remaining proportion of each non-to-be-added dish in the park canteen within the current prediction period; if the remaining proportion is higher than the preset second proportion threshold, then combined with the current dining data in the park canteen, the trained order prediction model is used to predict the remaining quantity of non-to-be-added dishes in the canteen at the end of dining in the park canteen, and generate the predicted remaining quantity of each non-to-be-added dish; for non-to-be-added dishes that are difficult to sell, it is necessary to predict their remaining quantity in advance at any time to facilitate timely processing.

[0032] When using, combine the contents in steps 401 to 403: After executing the menu arrangement plan, a menu ordering prediction model is established to predict the remaining quantity of dishes that are not to be added and dishes that are to be added. For dishes that are easy to sell and are to be added, when they are insufficient, they can be supplemented in appropriate amounts based on the insufficient quantity to fully guarantee the supply of dishes and improve the dining experience of employees in the park. For dishes that are difficult to sell and are not to be added, their possible remaining quantity is continuously predicted so that timely processing can be made to supplement the dishes that are in insufficient supply.

[0033] Step 5. Evaluate the dishes that may have surplus quantity and are not to be added, generate the dish evaluation coefficient Cpx, match the corresponding discount strategy for the dishes that are not to be added according to the dish evaluation coefficient Cpx, and choose the response method according to the execution effect of the discount strategy, and deal with the dishes that are not to be added and the dishes to be added separately on the next operating day; finally, based on the feedback, adjust the marketing strategies of both, reduce the operating risks under special weather conditions, and strike a balance between ensuring the meals of the employees in the park and its own operations.

[0034] The step five includes the following contents: Step 501: Obtain the predicted remaining quantity of each non-to-be-added dish, evaluate each non-to-be-added dish, and generate a dish evaluation coefficient Cpx; wherein the dish evaluation coefficient Cpx is obtained as follows: Get the dish price Cj and dish quantity Cv of each dish not to be added, as well as the sales volume of the dish on that day, and generate the dish sales volume Xw from the sales volume; after dimensionless processing of the above parameters, according to the following formula:

[0035] Among them, the parameters mean: weight coefficient, Price factor , is the food quantity factor, , is the sales factor, ; Among them, the method of obtaining the weight coefficient can refer to the hierarchical analysis method.

[0036] Obtain the dish evaluation coefficient Cpx, and match the corresponding discount strategy according to the distribution of the dish evaluation coefficient Cpx; adopt a larger discount for dishes that are relatively poor in quality and are not to be added, actively market the dishes that may be left over, and at the same time make up for the shortage of dishes to be added; For example, the canteen management terminal pre-sets a first evaluation threshold and a second evaluation threshold. When the dish evaluation coefficient Cpx is lower than the first evaluation threshold, the dish is given a first discount; when the dish evaluation coefficient Cpx is between the first threshold and the second threshold, the dish is given a second discount; when the dish evaluation coefficient Cpx is higher than the second evaluation threshold, the dish is given a third discount; Step 502: After obtaining the discount strategy generated by the dish evaluation coefficient Cpx, the discount strategy is implemented for dishes that are not to be added; after the current dining time is over, if there are still dishes that are not to be added remaining, the discount will be increased on the next operating day; if there are still dishes that are to be added remaining, the preparation quantity will be reduced.

[0037] When using, combine the contents in steps 501 to 502: After generating the dish evaluation coefficient Cpx and giving the corresponding discount strategy, a marketing strategy is formed. When there may be surplus food in the campus canteen, it can be dealt with in a timely manner. At the same time, some dishes can be supplemented to reduce operational risks. After the end of the day's operation, feedback can be obtained to adjust the marketing strategy.

[0038] See also Figure 2 The present invention provides an information intelligent management system based on artificial intelligence, comprising: The monitoring unit monitors the weather conditions when the canteen is in operation and the external weather is raining. The monitoring unit establishes a weather condition data set based on the monitoring results and obtains the rainfall coefficient Jxs. If the rainfall coefficient Jxs is higher than the condition threshold, an early warning message is sent to the canteen management end. The analysis unit, after receiving the warning information, uses the trained canteen dining model to predict the canteen transaction amount. If the predicted increase in the canteen transaction amount exceeds the expectation, a request message is sent out; if it does not exceed the expectation, the supply of some dishes is increased; After receiving the request information, the plan formulation unit obtains the historical dining records of the employees in the park cafeteria, and combines the dining data of the employees in each park, uses the trained dish allocation prediction model, predicts and generates the current dish arrangement plan of the park cafeteria, and sends the dish arrangement plan to the cafeteria management end; The prediction unit, after obtaining and executing the dish arrangement plan, uses the trained order prediction model to predict the remaining dishes in the current prediction period. If the remaining proportion of dishes to be added is lower than expected, a supplementary notification is issued to predict the predicted remaining quantity of each non-dish to be added at the end of the meal; The evaluation unit evaluates the dishes that may have surplus quantity and are not to be added, generates the dish evaluation coefficient Cpx, matches the corresponding discount strategy for the dishes that are not to be added according to the dish evaluation coefficient Cpx, and selects the response method according to the execution effect of the discount strategy, and handles the dishes that are not to be added and the dishes to be added separately within the next operation day.

[0039] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0040] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0041] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. An information intelligent management method based on artificial intelligence, Features: include, When the canteen in the park is in operation, if the external weather is in a rainy state, the weather conditions are monitored, and a weather condition data set is established based on the monitoring results to obtain the rainfall coefficient Jxs. If the rainfall coefficient Jxs is higher than the condition threshold, an early warning message is sent to the canteen management end; including: obtaining the current weather data, if it is currently in a continuous rainfall state, obtaining the current rainfall per hour, generating the average rainfall Pr, and obtaining the driving visibility Xd under the current conditions; when the average rainfall Pr exceeds the preset rainfall threshold, obtaining the traffic congestion data in the rainfall area, and identifying the average driving speed of the vehicle, comparing the speed with the speed under good weather conditions, and determining the ratio as the speed ratio Sd; summarizing the average rainfall Pr, driving visibility Xd and speed ratio Sd to establish a weather condition data set; After receiving the warning information, the canteen dining model after training is used to predict the canteen transaction amount. If the predicted increase in the canteen transaction amount exceeds the expectation, a request message is sent out; if it does not exceed the expectation, the supply of some dishes is increased; After receiving the request information, the historical dining records of employees in the park cafeteria are obtained, and combined with the dining data of employees in each park, the trained dish allocation prediction model is used to predict and generate the current dish arrangement plan of the park cafeteria, and the dish arrangement plan is sent to the cafeteria management end; After obtaining and executing the dish arrangement plan, the trained dish ordering prediction model is used to predict the remaining dishes in the current prediction period. If the remaining proportion of dishes to be added is lower than expected, a supplementary notification is issued to predict the predicted remaining quantity of each non-dish to be added at the end of the meal; Evaluate the dishes that may have surplus quantity and are not to be added, generate the dish evaluation coefficient Cpx, match the corresponding discount strategy for the dishes that are not to be added according to the dish evaluation coefficient Cpx, and choose the response method according to the execution effect of the discount strategy, and handle the dishes that are not to be added and the dishes to be added separately on the next operating day.

2. According to claim 1, an information intelligent management method based on artificial intelligence, Features: The rainfall coefficient Jxs is generated from the weather condition data set. The rainfall coefficient Jxs is generated as follows: the average rainfall Pr, driving visibility Xd and speed ratio Sd are dimensionlessly processed according to the following formula: Weight coefficient: , ,and ; If the rainfall coefficient Jxs is higher than the preset condition threshold, an early warning message will be sent to the canteen management end.

3. According to claim 1, an information intelligent management method based on artificial intelligence, Features: After receiving the early warning information, the canteen management end obtains weather data within the past year, obtains dining data in the park canteen, and separates the daily total transaction amount Ze of the park canteen from the dining data; summarizes the dining records of the park canteen and the corresponding weather data, and uses the neural convolution model to establish and generate a canteen dining model after training and testing; under the current rainfall conditions, uses the canteen dining model to predict the total transaction amount Ze and obtain the predicted transaction amount.

4. According to claim 3, an information intelligent management method based on artificial intelligence, Features: Get the total transaction amount Ze of the previous canteen operation day. If the predicted transaction amount is higher than the total transaction amount Ze, and the difference between the two is higher than the preset incremental threshold, send a request message to the outside; if it is not exceeded, get the top several dishes with the highest sales from the dining records of the canteen and determine them as dishes to be added; after determining the dishes to be added, ensure sufficient supply of dishes in the park canteen; after receiving the request message, increase the current supply of dishes to be added in the park canteen so that the increase in supply is consistent with the above difference.

5. According to claim 1, an information intelligent management method based on artificial intelligence, Features: After receiving the request information, the dining data of all park employees in the most recent recording period is queried from the dining records of the park cafeteria to form the historical dining records of employees; after receiving the warning information, the probability of each park employee going to the park cafeteria under the current weather conditions is obtained from the historical dining records to generate the dining probability; The dish preferences and dining probabilities of employees in each park are obtained from the employees' historical dining records, and a dining dataset is established after aggregation; a neural convolutional model is used to generate and output a dish allocation prediction model after training and testing using part of the data in the dining dataset; the current weather conditions, the dish preferences and dining probabilities of park employees are used as input, and the dish allocation prediction model is used to predict and analyze the dish arrangement of the park canteen on the day, and a dish arrangement plan for the park canteen is generated after iteration.

6. According to claim 5, an information intelligent management method based on artificial intelligence, Features: After receiving the menu arrangement plan and executing it, the canteen management end obtains the identity information of the employees entering the canteen and the current flow of people after the canteen enters the dining time period. The neural convolution model is used to generate a menu ordering prediction model after training and testing based on the above information, and several consecutive prediction cycles are set. During the current forecast period, the order prediction model is used to predict the remaining dishes in the campus cafeteria during the next forecast period and obtain the forecast results; if the remaining proportion of dishes to be added in the next forecast period is lower than the preset first proportion threshold, a dish replenishment notification is issued.

7. The information intelligent management method based on artificial intelligence according to claim 6, Features: When the number of people dining in the park canteen exceeds the preset threshold, the remaining proportion of each dish that is not to be added in the park canteen within the current prediction period is obtained. If the remaining proportion is higher than the preset second proportion threshold, the current dining data in the park canteen is combined with the trained ordering prediction model to predict the remaining amount of non-dishes to be added in the canteen at the end of the meal in the park canteen, and generate the predicted remaining amount of each dish that is not to be added.

8. The information intelligent management method based on artificial intelligence according to claim 7, Features: Obtain the predicted remaining quantity of each dish that is not to be added, evaluate each dish that is not to be added, and generate a dish evaluation coefficient Cpx; obtain the dish evaluation coefficient Cpx, and match the corresponding discount strategy according to the distribution of the dish evaluation coefficient Cpx; implement the discount strategy for the dishes that are not to be added; after the current dining time is over, if there are still remaining dishes that are not to be added, increase the discount on the next operating day; if there are still remaining dishes to be added, reduce the preparation quantity.

9. The information intelligent management method based on artificial intelligence according to claim 8, Features: The dish evaluation coefficient Cpx is obtained as follows: the dish price Cj and dish quantity Cv of each dish not to be added, as well as the sales volume of the dish on that day are obtained, and the dish sales volume Xw is generated from the sales volume; after dimensionless processing of the above parameters, according to the following formula: The parameters have the following meanings: is the price factor, , is the food quantity factor, , is the sales factor, .

10. An information intelligent management system based on artificial intelligence, Features: include: The monitoring unit monitors the weather conditions when the canteen is in operation and the external weather is raining. The monitoring unit establishes a weather condition data set based on the monitoring results and obtains the rainfall coefficient Jxs. If the rainfall coefficient Jxs is higher than the condition threshold, an early warning message is sent to the canteen management end. The analysis unit, after receiving the warning information, uses the trained canteen dining model to predict the canteen transaction amount. If the predicted increase in the canteen transaction amount exceeds the expectation, a request message is sent out; if it does not exceed the expectation, the supply of some dishes is increased; After receiving the request information, the plan formulation unit obtains the historical dining records of the employees in the park canteen, and combines the dining data of the employees in each park, uses the trained dish allocation prediction model, predicts and generates the current dish arrangement plan of the park canteen, and sends the dish arrangement plan to the canteen management end; After obtaining and executing the dish arrangement plan, the prediction unit uses the trained order prediction model to predict the remaining dishes in the current prediction period. If the remaining proportion of dishes to be added is lower than expected, a supplementary notification is issued to predict the predicted remaining quantity of each non-dish to be added at the end of the meal; The evaluation unit evaluates the dishes that may have surplus quantity and are not to be added, generates the dish evaluation coefficient Cpx, matches the corresponding discount strategy for the dishes that are not to be added according to the dish evaluation coefficient Cpx, and selects the response method according to the execution effect of the discount strategy, and handles the dishes that are not to be added and the dishes to be added separately on the next operation day.

Images