JPS62227377A - Wide area refuge induction support system - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention facilitates wide-area evacuation by estimating the reasons for refusal that generally exist in a specific area when there are many people who refuse wide-area evacuation. This relates to a wide-area evacuation guidance support system that cannot be advanced.

[Conventional technology]

There have been many evacuation behavior models for wide-area evacuation. However, all of these models are based on the perspective of when people would obey an evacuation order. Also,
Until now, there has been no wide-area evacuation guidance system that supports the smooth progress of wide-area evacuation based on the model described above.

[Problem that the invention seeks to solve]

The wide-area evacuation behavior model in the conventional wide-area evacuation guidance support system cannot estimate the reasons for refusal of those who refuse evacuation, which is always a problem during wide-area evacuation.

Therefore, there were problems in that appropriate persuasion was not carried out to those who refused, and the progress of the wide-area evacuation as a whole was often delayed.

The present invention has been made to solve the above-mentioned problems, and provides wide-area evacuation guidance support that estimates complex evacuation refusal reasons in a short time with simple operations when time is tight for wide-area evacuation. The purpose is to obtain a system.

[Failure to solve the problem]

The wide-area evacuation guidance support system according to the present invention stores a model that organizes and structures the reasons why people generally refuse to evacuate and information about a specific area, and allows the user to receive information specific to a particular disaster in an interactive format. It was designed to be acquired.

[For production]

The wide area evacuation guidance support system in the present invention extracts all applicable evacuation refusal reason candidates from the model based on the information and presents them to the user.

〔Example〕

Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

Figure 1 shows a system configuration diagram. In the figure, reference numeral 1 denotes a knowledge base, which is made up of a model unit 2 that stores models related to reasons for evacuation refusal, and a data unit 3 that stores information related to specific regional busy schedules. 4 is a user interface that exchanges information with the user; 5 is an inference unit that estimates the reason for evacuation refusal from a model based on the information in data Φ unit 3 and information from the user; 6 is a diagnostic result This is an explanation unit that is presented according to the user's request.

Next, the operation will be explained. First, the model unit 2 shown in FIG. 1 consists of a group of rules based on a model regarding reasons for refusal of evacuation, such as shown in FIG. 2, for example. The model consists of four main parts. That is, (A
) have no intention to evacuate, (f) cannot decide whether to evacuate, (e) have intention to evacuate but cannot evacuate,
(D) I have the intention to evacuate and am currently preparing. Figure 2 shows a more detailed classification of these four categories. The “” mark next to a valley reason indicates that the reason has a non-exclusive relationship with other reasons at the same level under the same higher-level reason (same time); other reasons are exclusive. Represents being in a relationship. The relationship between higher-level reasons and lower-level reasons in a model is considered to be a condition for the latter to hold true for the former. Model unit 2 expresses and stores such condition-reason pairs in production-rule format.

The reasons at the lowest level in FIG. 2 are specifically considered based on the information stored in the data unit regarding the specific region or disaster. All information is expressed in an attribute list format with object names, attribute names, and attribute values as variables. The attribute list contains
For example, there are object name:=area, attribute name:=name, attribute value:=area A, and so on. If the attribute list specified by the lowest reason exists in the data unit, that reason is true; if it does not exist, it is not true.

If the model cannot be considered based on data unit information alone, the unit 5 inferentially queries the user through the user interface unit 4. The user's response is simply a numerical selection from multiple options. This can reduce the burden on the user in an emergency.

The user's answers are converted into a property list, added to data unit 3, and used for subsequent diagnosis.

Next, the inference method in the inference unit 5 will be described. In wide-area evacuations, there are often multiple reasons for refusal. Therefore, inference unit 5 comprehensively examines the model from the top level and extracts all rejection reason candidates (ST-1 to ST-5).
T-8). At this time, inference is made backwards as shown in FIG. For example, if a reason at a certain level is Q, and when the reason ratio that has a non-exclusive relationship with Q is a non-exclusive condition of a higher-order reason P, in order to consider the establishment of the higher-order reason P, there is a level of 1. Consider reason Q. However, at this stage, it is unclear whether a reason Q at a certain level holds true, so we examine whether a reason at a level lower than that Q holds true (ST-9). Through a similar procedure, if the lowest reason is established based on the information in data unit 3, then the higher level reason P is established. Here, the superposed reason P has a non-exclusive relationship R with the reason Q at a certain level, so after the superordinate reason P is established by the reason Q at a certain level, the establishment of P by it is continued in the same way. Consider it in the procedure. If the above-mentioned Q and 几 are in an exclusive relationship, only one of Q and Bo will hold, so after the higher-order reason P is satisfied due to a reason Q at a certain level, Q
We do not examine reasons at a certain level that have a non-exclusive relationship with P, but we continue our reasoning by climbing backwards to the higher-level reason P. All established reasons are stored as evacuation refusal reason candidates (5T-4).

Next, explanation unit 6 will be explained. After the inference is completed, a refusal reason model consisting only of the stored evacuation refusal reason candidates is constructed. This model is a bare set of specific rejection reason candidates and lower-level rejection reason candidates. In explanation unit 6, explanations will be made based on this rejection reason model. As for the inference result, first, only the highest rejection reason is displayed as shown in FIG. “・” indicates a rejection reason candidate. If the user wants to know more detailed reasons, specify the corresponding number in the diagram. As a result, a list of the specified rejection reason candidate and the rejection reason candidates under the specified rejection reason candidate is displayed (FIG. 5). By repeating the same operations thereafter, it is possible to know down to the lowest level rejection reason candidates shown in FIG. However, the explanation can end at any level. After completion, the display shown in FIG. 4 is displayed once, and the system waits for input.

〔Effect of the invention〕

As described above, according to the present invention, special wide-area reasons for refusal of evacuation in various areas in various disasters can be handled under a unified framework, reducing development costs and time, and knowledge of reasons for refusal of evacuation. This has the effect of promoting systematization of the system.

[Brief explanation of drawings]

FIG. 1 is a system configuration diagram according to an embodiment of the present invention, FIG. 2 is a model diagram of the reason for evacuation refusal, FIG. 3 is an inference flowchart of the inference unit, and FIG. 4 and N! , 5 is an explanatory diagram of the inference results. In Figure 9, 1 is the knowledge base, 2 is the model unit, 3 is the data unit, 4 is the user interface unit, 5 is the inference unit, 6
is an explanation unit. Patent applicant: Mitsubishi Electric Corporation - - - - 1 Agent: Patent attorney 1) Hiroshi Sawa ゛, 1 (2 others) No intention to evacuate I; I/I; I/ 1, Indication of the incident: Akira Sawa No. 61-71472 3. Representative of the person making the amendment: Moriya Shiki (5) Date of amendment order: May 27, 1986 (7) Contents of the amendment (Attachment) Figure 2 will be amended. ni (: name 1 te Yusara cost 8
, One or more documents containing the revised Figure 2 of the attached documents

Claims (1)

[Claims] When a large number of people refuse to evacuate in a specific area when a wide-area evacuation order is issued during a disaster, etc., a model of the reasons for refusing evacuation is memorized to estimate the reasons why ordinary people would refuse to evacuate to that specific area. a knowledge base consisting of a model unit and a data unit storing various information regarding the specific area; and an inference unit that estimates the reason for evacuation refusal from the model based on the information in the data unit and information from the user. , the inference unit
A wide area evacuation guidance support system comprising an explanation unit that specifies the reason for evacuation refusal while exchanging information with an evacuation manager through an interface unit, and displays a diagnosis result in accordance with a user request.