Dayoub Y. Methods and models for optimization of transferring dangerous cargo

Українська версія

Thesis for the degree of Candidate of Sciences (CSc)

State registration number

0412U001396

Applicant for

Specialization

  • 01.05.04 - Системний аналіз і теорія оптимальних рішень

04-01-2012

Specialized Academic Board

Д 64.052.01

Kharkiv National University Of Radio Electronics

Essay

Subject of research - process of dangerous cargo transportation with accounting of emergency possibility. Aim of research - development of methods and models for optimizing transportation of dangerous cargo, targeted for operational, tactical and strategic decisions and decreasing losses and risks from emergency on current routes of transporting emergency cargo. Methods of research - system and statistical analysis for dynamic transporting task, methods of expert estimations, mathematical analysis, heuristic and mathematical programming for ant colony algorithms, multiple criteria optimization and decision support theory for guided evolution strategy. Hardware - personal computer. Theoretical and practical result of research is solution of scientific and technical task of developing methods and models for optimized dangerous cargo transportation. Obtained results have important scientific and practical task for creating systems, targeted on operative, tactical and strategic decisions. Developed methods and models allow to reduce risks and losses from emergency, produce safest routes in case of emergency, educate managers to control dangerous cargos transportations and produce recommendations for risks reduction. New scientific results include first formulation of dynamic transportation control task in case of forced change of planned route in case of emergency on current route; in order to solve proposed task a representation of transporting network as a graph of targeted evolution is proposed, model of decision support in strategic and operative solutions by using set of prior microsituation, which allowed to implement prognosis of future scenarios and reduce time for taking tactical or operative decisions on route change. For the first time method of operative search for control decisions is proposed in order to simplify route change process, which on the first step takes into account history of similar emergency situations as a set of prior microsituations, which implements on the next stop actively maintained evolution strategy with directed evolution mixed with ant-based colony model, which allowed to perform decision support in terms of more effective operative solutions on dangerous cargo transportation in minimal time. A new proposed method of multiple criteria optimization of dangerous cargo transportation route, in difference from previous solutions, uses successful previous solutions and uses in model of transport network heuristic generations of new alternate solutions and by adding fees on ineffective fragments allows optimization of strategic solutions of dangerous cargo transportation and allows to forecast risks values. A new method of risks calculation proposed in the paper for emergency risks calculation during dangerous cargo transportations, which, in difference from existing solutions, proposes reuse of generalized data on cargo transportation rates, on next step - configuration of transport network, geo-information data which allows to calculate safe speed of movement, amount of dangerous cargos and probability of cascade extensions of emergency situations. Results of the paper, especially methods of planning new optimal route with taking into account risks, are used during planning routes and performing training of routing teams in public joint-stock company "Eastern Ukrainian bank "Grant" (№ 2050/06-1-01 from 18.01.11) and KF AKB "TAS-CommerceBank" (№0411-01 from 04.11.10). Scientific theoretical and practical results of paper may be used in Ukrainian companies to form effective decisions in terms of multiple criteria, in educational process during studying process in informational technologies sphere.

Files

Similar theses