An algorithm for machine learning of a transport type model is presented for the optimal distribution of tasks in a heterogeneous group of robots operating in an automatic mode without operator participation. It is assumed that the model is trained by an experienced operator in a landfill environment adequate to a real emergency situation in which robots are to perform operations. According to the configured model in a real setting, tasks can be distributed according to a supervisory or decentralized control scheme. Training can be carried out and in the process of the regular operation of robots. In this case, the use of the learning model allows you to split the configuration tuning circuits and the assignment of tasks, which enables the robots and the operator to function at their own natural pace.
ÐÅÇÞÌÅ Ââåäåíèå. Ñèñòåìû ïîääåðaeêè ïðèíÿòèÿ ðåøåíèé ïðè óïðàâëåíèè ïîaeàðîòóøåíèåì ïîçâîëÿþò ñíèçèòü ïðÿìîé ìàòåðèàëüíûé óùåðá, êîëè÷åñòâî ïîãèáøèõ è ïîñòðàäàâøèõ. Ðàáîòà ïîñâÿùåíà ïîñòðîåíèþ ìîäåëè ðèñêà, çàêëþ÷åííîãî â ðåøåíèÿõ, ïðèíÿòûõ ðóêîâîäèòåëåì òóøåíèÿ ïîaeàðà (ÐÒÏ) êàê ëèöîì, ïðèíèìàþùèì ðåøåíèÿ (ËÏÐ), â êîíòåêñòå óïðàâëåíèÿ òóøåíèåì ïîaeàðà â ìíîãîýòàaeíîì çäàíèè. Öåëè è çàäà÷è. Öåëüþ èññëåäîâàíèÿ ÿâëÿåòñÿ ïîñòðîåíèå ìîäåëè, îòðàaeàþùåé óðîâåíü ðèñêà â ðåøåíèÿõ, ïðèíèìàåìûõ ÐÒÏ. Äëÿ äîñòèaeåíèÿ öåëè íåîáõîäèìî ðåøèòü ñëåäóþùèå çàäà÷è: 1) âûáðàòü òèï ìîäåëè ïðèíÿòèÿ ðåøåíèé; 2) ïîñòðîèòü àëãîðèòì îöåíèâàíèÿ ïàðàìåòðîâ ìîäåëè ïî íàáëþäåíèÿì çà ðåøåíèÿìè, ïðèíÿòûìè ÐÒÏ; 3) ïðîàíàëèçèðîâàòü êà÷åñòâî ìîäåëè. Ìåòîäû. Âûáðàí êëàññ ìîäåëåé ïðèíÿòèÿ ðåøåíèé, íàçûâàåìûõ èãðàìè ñ ïðèðîäîé. Ïðîöåäóðà ïðèíÿòèÿ ðåøåíèé ËÏÐ ïðè òóøåíèè ïîaeàðà â ìíîãîýòàaeíîì çäàíèè ïðåäñòàâëåíà â âèäå òðåõóðîâíåâîãî äåðåâà ðåøåíèé. Îíî ïðåîáðàçîâàíî â íîðìàëüíóþ (òàáëè÷íóþ) ôîðìó, ÷òî ïîçâîëèëî ïðåäñòàâèòü âûáîð ðåøåíèé â âèäå êðèòåðèÿ Ãóðâèöà. Ïàðàìåòð ïåññèìèçìà-îïòèìèçìà Ãóðâèöà îòðàaeàåò ñòåïåíü ðèñêà â ðåøåíèÿõ ËÏÐ. Äëÿ ïðîâåðêè ðàáîòîñïîñîáíîñòè ïðåäëîaeåííîé òåõíîëîãèè îöåíèâàíèÿ ïàðàìåòðà Ãóðâèöà âûïîëíåíî èìèòàöèîííîå ìîäåëèðîâàíèå. Ðåçóëüòàòû è èõ îáñóaeäåíèå. Èìèòàöèîííûå ýêñïåðèìåíòû ïîäòâåðäèëè ðàáîòîñïîñîáíîñòü ïðåäëîaeåííîé òåõíîëîãèè îöåíèâàíèÿ ñòåïåíè ñêëîííîñòè ËÏÐ ê ðèñêó. Îöåíêè, ïîñòðîåííûå äëÿ ðàçíûõ ËÏÐ, ïîçâîëÿþò ñðàâíèâàòü ñòåïåíü ñêëîííîñòè ê ðèñêó ðàçëè÷íûõ ÐÒÏ. Ýòî äàåò âîçìîaeíîñòü ïîñòðîèòü ýòàëîííûå îöåíêè ïî äàííûì î ïðèíÿòèè ðåøåíèé îïûòíûìè ÐÒÏ. Îöåíêè äðóãèõ ÐÒÏ ìîaeíî ñðàâíèâàòü ñ ýòàëîííûìè è äåëàòü âûâîäû î êà÷åñòâå èõ óïðàâëåíèÿ. Âûâîäû. Öåëü èññëåäîâàíèÿ äîñòèãíóòà ïóòåì ðåøåíèÿ ïîñòàâëåííûõ çàäà÷. Ïðåäëîaeåííàÿ òåõíîëîãèÿ ÿâëÿåòñÿ ðàçíîâèäíîñòüþ ìàøèííîãî îáó÷åíèÿ, ïðåäñòàâëÿåòñÿ ïåðñïåêòèâíîé è ìîaeåò áûòü èñïîëüçîâàíà â ñîñòàâå ñèñòåì ïîääåðaeêè ïðèíÿòèÿ ðåøåíèé ÐÒÏ, à òàêaeå ïðè îáó÷åíèè è ïîäãîòîâêå ïåðñîíàëà, çàíèìàþùåãîñÿ óïðàâëåíèåì â ÷ðåçâû÷àéíûõ ñèòóàöèÿõ. Êëþ÷åâûå ñëîâà: äåðåâî ðåøåíèé; èãðû ñ ïðèðîäîé; êðèòåðèé Ãóðâèöà; ëèöî, ïðèíèìàþùåå ðåøåíèÿ; îáðàòíàÿ çàäà÷à; îöåíèâàíèå. Äëÿ öèòèðîâàíèÿ: Âèëèñîâ Â. ß. Ìîäåëèðîâàíèå óðîâíÿ ðèñêà ðåøåíèé, ïðèíèìàåìûõ ïðè óïðàâëåíèè ëèêâèäàöèåé ïîaeàðîâ // Ïîaeàðîâçðûâîáåçîïàñíîñòü/Fire and Explosion Safety.-2019.-Ò. 28, ¹ 3.-Ñ. 36-49.
No abstract
The monograph examines topical issues of decision support and management in safety systems for fire and emergency situations through the use of innovative approaches and tools for operations research, artificial intelligence, robotics and management methods in organizational systems. The monograph is intended for faculty, researchers, graduate students (adjuncts) and doctoral students, as well as for undergraduates, students and listeners of educational organizations, all those who are interested in the problems of decision support and management in security systems.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.