Speed ​​prediction device and method thereof

Abstract

The present disclosure relates to a speed prediction device and method thereof. The speed prediction device according to this embodiment may include: a selection circuit for selecting a predicted section of a road for which a vehicle's speed is to be predicted, and a subsequent section located behind the predicted section in terms of the vehicle's traveling direction; a receiving circuit for receiving driving information including the speed of a vehicle passing through the predicted section and the subsequent section; a processing circuit for generating processed information including an average speed, the average speed being obtained by sequentially weighting the predicted section and the subsequent section in order of relative distance from the predicted section and calculating an average using the weights and the driving information; and a generation circuit for generating a predicted speed by inputting the processed information into a prediction model trained to predict the future speed of the vehicle.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to Korean Patent Application No. 10-2021-0175269, filed on December 9, 2021, which is incorporated herein by reference in its entirety. Technical Field

[0003] The present disclosure relates to a speed prediction device and method. Background Art

[0004] Conventionally, the arrival time of a vehicle at a destination is determined by collecting actual traffic data and predicting the arrival time directly from the collected data, or by predicting the arrival time through a mathematical model or simulation. However, these methods have difficulty reflecting changing traffic conditions in real time; thus, users who use guidance services derived from traffic information have to experience inconvenience.

[0005] Furthermore, it is assumed that road segments are assigned the same weight for estimating vehicle travel speeds within a specific road segment. In this case, due to the intersections (JCs), interchanges (ICs), and crossroads within a specific road segment, actual traffic conditions may not be accurately reflected for road segments with heavy traffic congestion or smooth movement. Consequently, it is difficult to accurately predict travel speeds and arrival times at destinations.

[0006] Recently, as artificial intelligence (AI) technology has received significant attention, a large amount of research on AI technology is being actively conducted; if technology related to artificial neural networks is used in addition to technology related to AI, it is expected that traffic conditions that change in real time can be predicted relatively accurately, and reliable information about the estimated traveling speed and arrival time of a vehicle can be provided to users.

[0007] The discussion in this section provides background information only and does not constitute an admission of prior art. Summary of the Invention

[0008] Based on the above background, according to one aspect of the present disclosure, an object of the present embodiment is to provide a technology for predicting the future traveling speed of a vehicle by selecting a prediction target section on a road and a rear section located behind the vehicle with respect to the vehicle's traveling direction, calculating the vehicle's traveling speed by weighting each road section differently, and training an artificial neural network using the calculated traveling speeds.

[0009] To achieve the above objectives, on the one hand, the present disclosure provides a speed prediction device, which includes: a selection circuit that selects a prediction section of a road for which the speed of a vehicle is to be predicted and a rear section located behind the prediction section in terms of the vehicle's travel direction; a receiving circuit that receives driving information including the speed of the vehicle passing through the prediction section and the rear section; a processing circuit that generates processing information including an average speed, which is obtained by weighting the prediction section and the rear section in sequence in order of relative distance from the prediction section and calculating an average value using the weights and the driving information; and a generation circuit that generates a predicted speed by inputting the processing information into a prediction model trained to predict the future speed of the vehicle.

[0010] To achieve the above object, according to another aspect, the present disclosure provides a speed prediction method, comprising: selecting a predicted section of a road for which a speed of a vehicle is to be predicted and a rear section located behind the predicted section in terms of a traveling direction of the vehicle; receiving travel information including a speed of a vehicle passing through the predicted section and the rear section, receiving time information including one or more pieces of information about a time, a date, a day of the week, and a month when the vehicle travels, and receiving section information including a length and a speed limit of each of the predicted section and the rear section; generating processed information including an average speed by weighting the predicted section and the rear section in order of relative distance from the predicted section and calculating the average speed. Obtained by calculating an average value using the weight and driving information; generating indicator information indicating the degree of road congestion in the following manner: calculating the actual travel time of the vehicle for each of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the speed value of the vehicle included in the driving information; calculating the travel time limit of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the corresponding speed limits; and dividing the actual travel time by the travel time limit; and generating a future predicted speed by calculating processing information, time information and indicator information and inputting the calculated information into a prediction model trained to predict the future speed of the vehicle.

[0011] As described above, the present embodiment provides an advantageous effect of relatively accurately predicting the vehicle's travel speed and arrival time to a destination by taking into account a specific road section and those sections located behind the specific road section. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 A speed prediction device according to this embodiment is shown.

[0013] Figure 2 A speed prediction method according to this embodiment is shown.

[0014] Figure 3: is a diagram showing selection of a predicted section and a subsequent section and calculation of a weighted average value of vehicle speed according to the present embodiment.

[0015] Figure 4 FIG. 1 shows a case where time information according to this embodiment is preprocessed into a vector form.

[0016] Figure 5 is a diagram illustrating input of processing information, time information, and indicator information into a prediction model according to the present embodiment. DETAILED DESCRIPTION

[0017] Hereinafter, some embodiments will be described in detail with reference to the illustrative drawings. When assigning reference numerals to the constituent elements of the respective drawings, it should be noted that the same constituent elements are intended to be assigned the same numerals as much as possible, even if shown in different drawings. In addition, when describing the present disclosure, if it is determined that a detailed description of a related well-known configuration or function incorporated herein unnecessarily obscures the main idea of ​​the present disclosure, the detailed description thereof will be omitted.

[0018] In addition, when describing the components of the present disclosure, terms such as first, second, A, B, (a), and (b) may be used. Such terms are intended only to distinguish one component from other components and do not limit the nature, order, or sequence of the components. If a component is referred to as being "coupled to," "in conjunction with," or "connected to" different components, it should be understood that the component is coupled to or connected to the different components, but another component may be "coupled," "in conjunction with," or "connected" between the two components.

[0019] Figure 1 A speed prediction device according to this embodiment is shown.

[0020] The speed prediction apparatus 100 according to this embodiment may include a selection circuit 110 , a reception circuit 120 , a processing circuit 130 , a generation circuit 150 , and a learning circuit 140 .

[0021] The selection circuit 110 can select the predicted road section for predicting the speed of the vehicle and the rear road section behind the predicted road section with respect to the direction of travel of the vehicle. Here, the rear road section is behind the predicted road section with respect to the direction of travel of the vehicle and can include one or more continuous road sections.

[0022] The receiving circuit 120 can receive travel information including the speed of the vehicle passing through the predicted road section and the subsequent road section selected by the selection circuit 110. Furthermore, the receiving circuit 120 can receive time information including one or more of the time, date, day of the week, and month of the year when the vehicle traveled; and road section information including the length and speed limit of each of the predicted road section and the subsequent road section. The receiving circuit 120 can collect each piece of information using a detection device installed in each vehicle or using a detection device installed on the road.

[0023] Here, the section information may be received through the receiving circuit 120 as described above, and the section information regarding the predicted section and the subsequent section on the road may be pre-stored in the speed prediction apparatus 100 .

[0024] The processing circuit 130 may generate processing information including an average speed obtained by sequentially weighting the predicted road section and the subsequent road section in order of relative distance from the predicted road section and calculating an average value using the weights and the driving information. Figure 3 Describes in detail how to calculate average speed using weights.

[0025] The processing circuit 130 may also preprocess the driving information received by the receiving circuit 120 into a matrix format. In this case, the matrix containing the preprocessed driving information may be used to set the vehicle's driving speed based on the passage of time and traffic flow on the road. Furthermore, time information received by the receiving circuit 120, including at least one or more pieces of information regarding the time, date, day of the week, and month of the year during which the vehicle was traveling, may be preprocessed into a vector format.

[0026] The learning circuit 140 can input the processing information into the prediction model so that the prediction model learns the processing information. Here, the prediction model can use an artificial neural network model, and a deep neural network model can be used for the artificial neural network model.

[0027] The learning circuit 140 can train the prediction model by inputting processed information including the average speed of the vehicle into the prediction model along with time information, indicator information, or both. Prior to inputting the information into the prediction model, the processed information, time information, and indicator information can be intercalated, and the corresponding calculation can be a multiplication operation. Because the prediction model learns the vehicle's past speeds in the predicted section and subsequent sections, as well as the time, date, day of the week, and indicators indicating delays in the predicted section and subsequent sections, the degree of delay and vehicle speed in the predicted section can be predicted by providing the trained prediction model with the real-time speed of the vehicle traveling on the road, the current time, date, day of the week, and the real-time delay indicator.

[0028] The speed prediction apparatus 100 may further include an indicator generating circuit (not shown).

[0029] The indicator generation circuit can generate indicator information representing the degree of road congestion in the following manner: calculate the actual travel time of the vehicle for each of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the speed value of the vehicle included in the driving information, calculate the travel time limit of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the corresponding speed limit, and divide the actual travel time by the travel time limit.

[0030] The indicator information may be generated by dividing the travel time limit by the average travel time or by dividing the average travel time by the travel time limit.

[0031] For example, suppose the predicted section is section A, the subsequent section is section B, the length of sections A and B is 1000m, the speed limit of section A is 10m / s, the speed limit of section B is 20m / s, and the speed of vehicles passing through sections A and B is 10m / s. Considering that despite the speed limit of section B being 20m / s, vehicles actually travel at 10m / s in section B, it can be inferred that section B is relatively congested compared to section A. When using the length of each section and the travel time of vehicles for each section, the actual travel time for section A may be 100s, and the travel time limit for section A may be 100s. The actual travel time for section B may be 100s, and the travel time limit for section B may be 50s. Therefore, if the actual travel time is divided by the travel time limit to calculate the indicators for road section A and road section B, the indicator for road section A may be 1, and the indicator for road section B may be 2. Therefore, it can be seen that if the indicator is 1, the traffic condition of the corresponding road section is smooth, and if the indicator is higher than 1, the traffic condition of the corresponding road section is delayed or congested.

[0032] In addition, the indicator generation circuit may pre-process the indicator information and the road section information calculated using the processed information into a matrix form. Here, in the matrix including the pre-processed indicator information, an indicator representing the degree of road traffic congestion may be set according to the passage of time and traffic flow on the road.

[0033] The generation circuit 150 may generate a predicted speed by inputting the processing information into the trained prediction model generated by the learning circuit 140. The generation circuit 150 may input the processing information into the trained prediction model together with the time information, the indicator information, or both the time information and the indicator information. Before being input into the trained prediction model, the processing information, the time information, and the indicator information may be mutually calculated, where the corresponding operation may be a multiplication operation.

[0034] The vehicle speed included in the driving information received by the receiving circuit 120 represents the vehicle speed for the past two hours from the current time point (real time), and the predicted speed generated by the generating circuit 150 may represent the vehicle speed for two hours from the current time point (real time). Here, the current time point may mean the real-time point based on which the user wants to obtain the future predicted speed through the speed prediction device.

[0035] The speed prediction device may further include a transmission circuit (not shown). The transmission circuit transmits information about the predicted speed to a route search server, which analyzes the predicted speed generated by the generation circuit 150 and generates shortest-time route information. The route search server may generate the shortest-time route information and transmit the generated information to the user's navigation device, so that the user can reach the destination in the shortest time by driving the vehicle according to the shortest-time route information.

[0036] Figure 2 A speed prediction method according to this embodiment is shown.

[0037] The speed prediction method according to this embodiment may perform step S210: selecting a predicted road segment for which the vehicle speed is to be predicted and a subsequent road segment located behind the predicted road segment in terms of the vehicle's travel direction. Here, the subsequent road segment is located behind the predicted road segment in terms of the vehicle's travel direction and may include one or more consecutive road segments.

[0038] Step S220 may be performed to receive driving information including the speed of the vehicle traveling through the predicted road segment and the subsequent road segment; receive time information including one or more pieces of information regarding the time, date, day of the week, and month of the year when the vehicle was traveling; and receive road segment information including the length and speed limit of each of the predicted road segment and the subsequent road segment. Each piece of information may be collected using a detection device installed in each vehicle or using a detection device installed on the road. A preprocessing step may also be performed to arrange the driving information in a matrix form. Furthermore, a preprocessing step may also be performed to arrange the time information in a vector form.

[0039] Step S230 may be performed: generating processing information including an average speed obtained by weighting the predicted section and the subsequent section in order of relative distance from the predicted section and calculating an average value using the weight and the driving information. Figure 3 Detailed Description The average speed is calculated based on a weighted average.

[0040] Step S240 may be performed: calculating the actual travel time of the vehicle for each of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the speed value of the vehicle included in the driving information, calculating the travel time limits of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the corresponding speed limits, and generating indicator information indicating the degree of road congestion by dividing the actual travel time by the travel time limit.

[0041] The indicator information can be generated by dividing the travel time limit by the actual travel time, or by dividing the actual travel time by the travel time limit. The indicator information can be preprocessed into a matrix format. Indicators representing the degree of road traffic congestion can be arranged in the matrix including the preprocessed indicator information based on the passage of time and traffic flow on the road.

[0042] Step S250 may be performed: calculating the processing information, indicator information, and time information in real time and inputting the calculated information into a prediction model trained to predict future vehicle speeds. Before inputting into the prediction model, the processing information, time information, and indicator information may be mutually calculated, and the corresponding operation may be a multiplication operation.

[0043] Step S260 may be executed: generating a predicted future speed. Here, the vehicle speed included in the received driving information represents the vehicle speed for the past two hours from the current time point (in real time), and the generated predicted speed may represent the vehicle speed for two hours from the current time point (in real time). Here, the current time point may refer to the real-time point at which the user desires to obtain a predicted future speed using a speed prediction device or method.

[0044] The speed prediction method according to this embodiment may further include the steps of calculating processing information, time information, and indicator information, and inputting the calculated information into a prediction model so that the prediction model learns the processing information, time information, and indicator information. The prediction model may use an artificial neural network model, and a deep neural network model may be used for the artificial neural network model. The processing information, indicator information, and time information used to train the prediction model may be information accumulated over at least the past year.

[0045] Figure 3 : is a diagram showing selection of a predicted section and a subsequent section and calculation of a weighted average value of vehicle speed according to the present embodiment.

[0046] The predicted road segment 310, the first rear road segment 320 and the second rear road segment 330 may be selected to predict the vehicle's travel speed. Figure 3A total of two rear sections are selected in , but one rear section may be selected, or three or more rear sections may be selected.

[0047] Weights can be applied sequentially to the predicted road segment and the subsequent road segments 320 and 330 to calculate an average speed for predicting the vehicle's travel speed in the predicted road segment 310. For example, assuming the vehicle speed in the predicted road segment 310 is 30 km / h, the vehicle speed in the first subsequent road segment 320 is 40 km / h, and the vehicle speed in the second subsequent road segment 330 is 50 km / h. A weight of 3 can then be input to the predicted road segment 310 for which the travel speed is to be predicted. A weight of 2 can be input to the first subsequent road segment 320 that is relatively close to the predicted road segment 310, and a weight of 1 can be input to the second subsequent road segment 330 that is relatively farther from the predicted road segment 310. Therefore, if the average speed is calculated using the vehicle speeds and weights in each of the sections 310, 320, and 330, the weighted average speed can be calculated by adding the speeds in the predicted section 310 with a weight of 3 three times, adding the speeds in the first rear section 320 with a weight of 2 twice, and adding the speeds in the second rear section 330 with a weight of 1 once, and then dividing the sum of the speeds by the sum of the weights. Therefore, if the average speed is calculated according to the above procedure, the average speed obtained is (30+30+30+40+40+50) / (3+2+1)=36.67 km / h. In other words, although the real-time driving speed of the vehicle in the predicted section 310 is 30 km / h, when the vehicle speeds in the first rear section 320 and the second rear section 330 are taken into account, the average speed of the vehicle can be calculated to be 36.67 km / h. After calculating the average speed of multiple vehicles passing through the prediction section 310, the first rear section 320, and the second rear section 330 as described above, the calculated average speed can be input to the prediction model for learning. Based on the above operation, the predicted speed can be derived by considering the prediction section 310 and the rear sections 320 and 330.

[0048] The above operation can be expressed as follows. Assume that there is a predicted segment and n-1 subsequent segments, and weights of n, n-1, n-2...1 are assigned to the predicted segment and the subsequent segments in the order of relative proximity to the predicted segment. Assume that V1, V2, V3...V n It means the speed of the vehicle passing the predicted section and the following section in the order of being relatively close to the predicted section. At this time, the average speed V of the vehicle is calculated considering the weight. wa The equation can be expressed as follows.

[0049] [Equation 1]

[0050]

[0051] As described above, the vehicle's travel speed in the predicted section 310 can be more accurately predicted by using the average speeds in the predicted section and the subsequent sections 320 and 330 and their weights.

[0052] Figure 4 FIG. 1 shows a case where time information according to this embodiment is preprocessed into a vector form.

[0053] The receiving circuit installed in the speed prediction device according to the present embodiment can receive time information including at least one or more pieces of information of the time, date, day of the week, and month when the vehicle is traveling.

[0054] like Figure 4 As shown, the time information can be preprocessed into a vector form. The processing circuit in the speed prediction device according to this embodiment can perform the preprocessing of the time information. Figure 4 As shown, the vector including time information may include information about the time when the vehicle travels, information about the day of the week when the vehicle travels, information about the date when the vehicle travels, and information about the month when the vehicle travels.

[0055] Specifically, if Figure 4 As shown, in the vector including time information, a time value corresponding to two hours elapsed from the current (real-time) time point of vehicle travel may have a specific value other than 0. In addition, information on the day, date, and month of vehicle travel may have a specific value other than 0. Figure 4 As shown, the value of the vector including time information may be 1, or may have a value other than 1 to assign weight to time and date.

[0056] Therefore, a vector including time information is calculated in association with driving information and indicator information, wherein the driving information includes the driving speed of the vehicle and the indicator information represents the degree of traffic congestion in the predicted road section and the subsequent road section, and is used to train a prediction model to predict the driving speed of the vehicle at a specific time, date, and day of the week.

[0057] Figure 5 is a diagram illustrating input of processing information, time information, and indicator information into a prediction model according to the present embodiment.

[0058] like Figure 5 As shown, input data 1 510, input data 2 520, and input data 3 530 may be input to the prediction model 540 according to the present embodiment. Here, input data 1 510 and input data 2 520 may be processed information indicating an average speed calculated based on a weighted average of the vehicle's traveling speeds received by the receiving circuit, and indicator information indicating the degree of traffic congestion on the predicted road section and the subsequent road section.

[0059] The indicator information may be generated by calculating the actual travel time of the vehicle for each of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the speed value of the vehicle included in the driving information, calculating the travel time limits of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by the corresponding speed limits, and dividing the actual travel time by the travel time limit. The indicator information may be generated by dividing the travel time limit by the actual travel time or by dividing the actual travel time by the travel time limit.

[0060] The processing circuit of the speed prediction device may preprocess the processed information into a matrix form, wherein the elements of the matrix are arranged according to the passage of time and traffic flow on the road. The indicator generation circuit of the speed prediction device may preprocess the indicator information into a matrix form, wherein the elements of the matrix are arranged according to the passage of time and traffic flow on the road.

[0061] Input data 3 530 may be time information. The time information may include one or more pieces of information about the time, date, day of the week, and month when the vehicle is traveling. The processing circuit of the speed prediction device may pre-process the time information into a vector form.

[0062] like Figure 5 As shown, since the processing information, indicator information and time information are input to the prediction model 540, the future vehicle travel speed can be predicted using the average vehicle speed information including the traffic congestion level on the road segment and the information about the time and date when the vehicle travels.

[0063] The prediction model 540 may use a deep neural network (DNN) model, and the input data 1 510 , the input data 2 520 , and the input data 3 530 may be applied to a calculation process with each other before being input to the prediction model 540 .

[0064] Unless otherwise expressly stated, since the terms "include," "comprise," or "have" used throughout this document mean that the corresponding constituent elements may be included in the embodiments, they should be interpreted as further including other constituent elements rather than being excluded. Unless otherwise specified, all terms (including technical or scientific terms) used in this disclosure provide the same meanings as those commonly understood by those skilled in the art to which the disclosure belongs. Those terms defined in ordinary dictionaries should be interpreted as having the same meanings as conveyed in the context of the relevant technology. Unless otherwise expressly defined in this disclosure, those terms should not be interpreted as having ideal or overly formal meanings.

[0065] The description given above is merely an embodiment for illustrating the technical principles of the present disclosure, and without departing from the inherent characteristics of the present disclosure, a person skilled in the art may make various changes and modifications to the present disclosure. Therefore, it should be understood that the embodiments disclosed in this specification are not intended to limit the technical principles of the present disclosure, but rather to support the description of the present disclosure, and therefore the technical scope of the present disclosure is not limited by the embodiments. The technical scope of the present disclosure should be judged by the appended claims, and all technical principles found within the scope equivalent to the technical scope of the present disclosure should be interpreted as belonging to the technical scope of the present disclosure.

Claims

1. A speed prediction device, comprising: a selection circuit that selects a predicted road section for a road whose speed of a vehicle is to be predicted and a rear road section located behind the predicted road section with respect to a traveling direction of the vehicle; a receiving circuit for receiving driving information including the speed of the vehicle passing through the predicted road section and the subsequent road section; a processing circuit that generates processing information including an average speed obtained by sequentially assigning weights to the predicted road sections in order of relative distance from the predicted road section and assigning higher weights to the subsequent road sections, and calculating an average using the weights and the travel information; generating circuitry to generate a predicted speed by inputting the processed information into a prediction model trained to predict a future speed of the vehicle; as well as an indicator generating circuit configured to generate indicator information indicating a road congestion degree of each of the predicted road section and the subsequent road section, The generating circuit generates the predicted speed by calculating the processing information and the indicator information and inputting the calculated information into the trained prediction model.

2. The device according to claim 1, wherein The receiving circuit receives time information including one or more pieces of information about the time, date, day of the week, and month when the vehicle is traveling, and The generating circuit generates the predicted speed by calculating the time information and the processing information and inputting the calculated information to the trained prediction model. 3 . The apparatus according to claim 1 , further comprising a learning circuit that inputs processing information accumulated from the past into the prediction model and causes the prediction model to learn the processing information.

4. The device according to claim 1, wherein The receiving circuit receives the section information including the length and speed limit of each of the predicted section and the subsequent section, and The indicator generation circuit generates indicator information indicating a degree of road congestion by: calculating an actual travel time for the vehicle on each of the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by a speed value of the vehicle included in the driving information; calculating a travel time limit for the predicted section and the subsequent section by dividing the lengths of the predicted section and the subsequent section by corresponding speed limits; and dividing the actual travel time by the travel time limit. 5 . The apparatus according to claim 1 , further comprising a transmitting circuit that transmits information about the predicted speed to a route search server, the route search server generating shortest time route information by analyzing the predicted speed.

6. The device according to claim 1, wherein The vehicle speed included in the travel information represents the vehicle speed for two hours past from a current time point, and the predicted speed represents the vehicle speed for two hours from the current time point.

7. The device according to claim 1, wherein The prediction model is a deep neural network model.

8. A speed prediction method, comprising: selecting a predicted road section for a road whose speed of a vehicle is to be predicted and a rear road section located behind the predicted road section with respect to a traveling direction of the vehicle; receiving travel information including a speed of the vehicle passing through the predicted road section and the subsequent road section, receiving time information including one or more pieces of information regarding a time, date, day of the week, and month when the vehicle traveled, and receiving road section information including a length and a speed limit of each of the predicted road section and the subsequent road section; generating processing information including an average speed obtained by assigning weights to the predicted road sections in order of relative distance from the predicted road sections and assigning higher weights to the subsequent road sections, and calculating an average using the weights and the travel information; generating indicator information representing a road congestion degree by: calculating an actual travel time of a vehicle for each of the predicted road section and the subsequent road section by dividing the lengths of the predicted road section and the subsequent road section by a speed value of the vehicle included in the travel information, calculating a travel time limit for the predicted road section and the subsequent road section by dividing the lengths of the predicted road section and the subsequent road section by corresponding speed limits, and dividing the actual travel time by the travel time limit; as well as A future predicted speed is generated by calculating the processing information, the time information, and the indicator information and inputting the calculated information into a prediction model trained to predict a future speed of the vehicle.

9. The method of claim 8, further comprising inputting the processing information, the time information, and the indicator information accumulated over at least the past year into the prediction model that learns input information.

10. The method according to claim 8, further comprising preprocessing the time information into a vector form.

Images