M.Sc Thesis

M.Sc StudentShechter Yativ
SubjectInfo-gap robustness analysis of the design of a heavy
weapon station with load uncertainty
DepartmentDepartment of Mechanical Engineering
Supervisor PROFESSOR EMERITUS Yakov Ben-Haim
Full Thesis text - in Hebrew Full thesis text - Hebrew Version


The weapon system is exposed to diverse and complex external forces. The load uncertainty is due to the variable forces from firing different types of ammunition, and external forces that occur while the vehicle rides on irregular and unfamiliar terrains. This load uncertainty affects the weapon displacement while firing, which directly affects the weapon system accuracy. In the design of such a weapon system the robustness of the design is one of the most important design parameters, because the urban and modern battlefield demands surgical and extremely accurate weapon systems.

In order to overcome those uncertainties different methods have been developed. Many studies approach this problem from different points of view in order to optimize and overcome the load uncertainty. In this work we review several methods related to decision making under severe uncertainties. We focus on probabilistic and non probabilistic methods, describe Robust Satisficing method and elaborate on info-gap decision theory.

In this work we demonstrate and develop a method to design a heavy weapon system in the robustness approach using the info-gap method.  Robustness analysis demands a system model so we developed a mechanical model for the weapon system. This model is composed of multiple point masses connected by mass less rods with torsion springs at the joints.  The design requirement constrains the maximum displacement of the last mass from steady state while the system is subjected to external forces with severe uncertainties.

In the work we apply three different info-gap models as part of the info-gap robustness analysis. The first is the uniform-bound model which constrains the uncertain deviation by a scalar number. This unknown scalar number is unbounded because the uncertainty is unbounded. The second is the energy-bound model which deviates in an unknown transient manner from the nominal value.  The third is the Fourier-Bound model, which is suggested in cases when the designer has partial spectral information for characterizing the uncertainty.

We also discuss the Value-of-Information. How the information on the force frequency can help the designer achieve much better robustness values. We illustrate the analysis by using the Fourier info-gap model and conclude a strong relation between the natural frequencies of the system and the force frequency. We suggest that the value of the firing frequency will be about a fourth of the first natural frequency value, which helps achieve up to 5 times the robustness.