Research On Capability Generation Oriented Weapon Network Portfolio Optimization

Capability-based planning(CBP)has become a crucial mission for many countries around the globe including the American,British,Canadian,and Australian Defense Forces in future force design which can operate effectively in a diverse range of future scenarios from a system of systems(SoS)perspective.An SoS is the collection of individual and independent systems that are brought together for specific goals.Each system contributes to the SoS with unique capabilities and the cumulative capabilities achieved from their interoperability.SoS architecting administers appropriate integration of the systems as the fundamental of CBP.The capability is defined as the capacity or ability to achieve an operational effect.An operational effect may be defined or described regarding the nature of the effect and of how,when,where,and for how long it is produced.The main task of CBP is to design a weapon architecture by defining intended planning capabilities to address future uncertain scenarios effectively.With the continuous development of information technology,the concept of “Network-Centric Warfare” has emerged as a new warfare pattern which integrates sensors,decision makers and weapon systems through information networks to achieve better battlefield synchronization effects.As a consequence,the weapons are more complexly interrelated,and the network model has become a basic way to describe a weapon SoS.To achieve competitive advantages in the battlefield,weapons collaborate with each other to form a network providing multi-capability products has become the key to defeat enemies.Therefore,the research of this paper focuses on two aspects including how to construct a network-based weapon SoS and how the capability is delivered and generated through the collaboration network.Around the problem of weapon network portfolio optimization,we conduct research on the network-based capability generation model with network theory and multiobjective optimization method.Facing the practical problems in different stages of weapon SoS architecting and planning,the study of this paper is divided into four steps,first starting with single-scenario capability generation modeling,then a many-objective optimization research is conducted on the multi-scenarios situation.In order to measure the countermeasure effectiveness caused by capability aggregation,the concept of capability chain and threat elimination model is proposed.Finally,from a long-term perspective,planning the most potential weapon development route is studied with a reinforcement learning method.These four steps are closely connected with each other.All four parts focus on the same problem of weapon network portfolio optimization,and the network-based capability generation model in the second part serves as the basis for SoS-level capability calculation in other parts.The main contribution of this paper is as follows:(1)The capability generation model about how the capability is delivered across the collaboration network is studied.First,according to the modern warfare features,the difference between the weapon systems portfolio and the traditional portfolio problem is analyzed.To meet diverse capability needs by architecting different types of systems into a network,we formulate and analyze an So S architecting problem with network theory as a multi-objective optimization model.More specifically,we construct a three-step capability delivery model from a local and global perspective to describe the positive and negative consequences caused by system interactions.A three-level response model with a feasibility concept is proposed as the conversion of the capability value to meet different scale combat needs and balance the deployment of new and old weapons.And then,certain intelligent optimization algorithms incorporated with the proposed model is designed to approximate Pareto fronts of three response levels respectively.Results from a set of numerical studies demonstrate that an efficient structure with fewer nodes performs better than a full set structure,and the deployment of a system should be a comprehensive consideration of its overall structural influence rather than single system performance.(2)Based on the first part,research is extended to a multi-scenarios situation.First,relative to single scenario,we analyze the characteristics of multi-scenarios problem.The capability requirements in different scenarios vary a lot,and the weapon sets are also shared in different scenarios with different performance levels.The differences in weapon sets of these scenarios also result in different collaboration patterns.Therefore,the values of a weapon network solution in these scenarios are often conflicting and must be evaluated to make trade-offs.As the number of scenarios increase,the general multiobjective optimization algorithm is no longer applicable.Many-objective optimization model based on scenario decomposition for weapon network portfolio in multiple scenarios is proposed.By analyzing the optimized results,the hierarchical structure of the weapon network is found,and the concepts of basic weapon network,general weapon package,and specific weapon are proposed.It also reflects the idea of So S warfare and modular weapon architecture design in the future.(3)The essence of the future warfare is the confrontation between capability packages.Focusing on how to evaluate the countermeasures effectiveness by capability aggregation,and how to guide the weapon development planning according to the achievement of the effect,we believe that the key to SoS breaking warfare is to gain superior in multi-dimensional capabilities and eliminate the threat of enemy forces.Since the capability is the functional description of weapon SoS,and the weapon has composite functions,the traditional operation loop and the kill chain theory cannot fully reflect the functional interaction among them.By analyzing the interrelations among capabilities at a higher level,this paper proposes the concept of capability chain,which aggregates the capability to a threat elimination model,and use it as an objective to optimize the future weapon planning..(4)From a long-term development perspective,we conduct research on the forwardlooking multi-stages weapon portfolio planning problem.In the traditional multi-stage planning research,the value at the current stage of the solution is always used as the basis of solution selection.However,the current solution often affects the future alternative solutions in rolling planning horizon.In order to explore the potential value of solutions and find a future-oriented weapon development planning route,which can make the overall score optimal,Inspired by the reinforcement learning ideas,this paper combines reinforcement learning with multi-objective optimization through exploring and exploiting mode,superimpose the possible future value into the evaluation of the current plan,and finally select the optimal plan of each stage by considering the value matrix of future value,and then get the optimal planning route.