The Problems And Approachs For Capabilities-based Armaments Portfolio Planning

The capstone planning of multiple armaments constructions is a complex system engineering, and it must to make an overall decision on the model selection, development time and budget for each kind of armament. Therefore, the capstone planning determines the development orientation, composition and structure, and the combatant level of multiple araments. However, there are several challeges exsiting in the capstone planning of the armaments. Firstly, with the development of defesne economy and military technology, it becomes larger on the quantities of the kind and model of the araments related to the capastone planning, which results in the huge alternative space and similar weapon models. Secondly, each kind of armaments is embraced with large development risks resulted from the application of high-tech and high investments, thus, multiple objectives, such as operatins capabilities, development risks and investement budegts, should be trade-off to reach a balance for the development of aramremts.Thirdly, with the complexity and diversity of the safety threates, the military missions and military demands are evolving dynamically. Thus, all of these lead to the high uncertainties for the objectivs of the capastone planning of the multiple armaments.Capabilties Based Planning(CBP) is a top-to-down capastone philosophy to instruct the multiple armaments from various armed forces and response the uncertainty chanlleges in an economic framework, it has been an important composition of the U.S. defense decion-making. However, CBP is lack of quantitatively and systemetic decion-making and optimization approaches to support the capastone planning of armaments more effectively. Therefore, this thesis proposes a research topic on Capabilities based Armaments Portfolio Planning(CAPP), which aims at the realistic problems in capastone planning of araments and refers the latest development concept i.e. Capabilites Portofolio Management(CPM). Furthermore, this thesis integrates related decision theroies and optimzaiton tecniques such as multiobjective decision-making, evolutionary algorithm and stochastic programming, and investigates the scientific probems and solution approachs in CAPP. The main results and contributions of this thesis are presented as follows.(1)The essential problem model, decision-making model and intelligence optimization algorithm are proposed for CAPP respectively. The capability requirements, planned capabilities and constraints are respectively modeled by analysis of the concept and definition on CAPP. The three kinds of essential decision-making model i.e. weight-, preemption- and robust- are proposed to solve the contradiction and immeasurability of the multiple capability requriements and the various perferneces of decision makers. A differential evolution based optimization algorithm is proposed to considering the huge solution space which consists of the variables of the model selection, development time and construction budget.(2)A complex multiobjective planning problem and the corresponded approach are proposed considering development risks of the planned capabilities of the armament portfolio. In order to meet multiple capability requirements, avoid the corresponed multiple risks and save the construction cost, a complex multiobjective planning approach framework is proposed. The framework applys the expected downsied risk measure to model the uncertainty of the capability development, and proposes the capability-risk and the capability-risk-cost multiobjective optimization model. An interactive multiobjective differential evolution algorithm is proposed to acquire the non-inferior set which belongs to the interest range to decision makers. Finally, a forcasting and optimization based ideal solution algorithm is developed to calclulate the compromised solution from the non-inferior set. Compared with normal ideal solution technique, the propsed algorithm can generate a compromise solution which can meet the preference of decision makers more accurately.(3) A multistage planning problem and corresponded stochastic programming approach are proposed for discrete and evolving capabitliy requirements. According to the capability requirements evolving to mapping with the change of the military missions in the whole planning horizon, a multistage programming problem model is proposed for CAPP. Aiming at the discrete uncertainty generated by evolving capability requirements, a multistage stochastic programming approach is proposed for CAPP. This approach firstly constructs a multistage deterministic programming model based on the goal programming technique. Then a scenario tree is constructed to quatitatively describe the discrete and evolving capability requirements, meanwhile, the mean based- and the robust based- stochastic programming model are proposed to cope with different risk preferences of the decision makers. Finally, considering the variable dimesions increases exponentially with the branches of the scenario tree, a variable clustering based parallel stochastic differential evolution algorithm is proposed to solve the stochastic programming model.(4)A multistage on-line programming approach is proposed for interval based evolving capability requirements. When the possible values of the evolving capability requirements exist as an interval numbers, the off-line stochastic programming approach is hard to meet the interval based uncertain capability requirements accurately. To solve this chanllege, a multistage on-line programming approach is proposed. This approach only determines the model selection plan of the armament portfolio at the initial point of the decision-making through using the multistage stochastic programming and the variable clustering principle. Then, this approach imports the on-line optimization concept to construct the on-line stochastic programming model and solve this model on the decision making point in each planning stage, in order to generate the portfolio plan on the development time and consturcuion budget for the armaments portfolio.