Traditional Sensitivity Analysis Research Articles

Verification and Validation of Hierarchical Fault Diagnosis in Satellites Formation Flight

It is well known that for long-duration space missions, there is growing need for efficient utilization of telemetry data to enhance diagnostic performance and assist the less-experienced personnel in performing monitoring and diagnosis tasks. To address this need, we have, recently, developed a systematic and transparent fault diagnosis methodology within a hierarchical fault diagnosis framework for satellites formation flight. We developed our proposed hierarchical decomposition framework through a novel Bayesian network-based model, namely component dependence model (CDM). In this paper, we investigate the verification and validation of the CDM for fault diagnosis in satellites formation flight. We propose and develop a sensitivity analysis to verify the CDM by taking advantage of our systematic CDM development methodology. The proposed verification method satisfies the unique requirement of identifying CDM sensitivity when diagnostic performances of the algorithms that are deployed at one or more nodes of the CDM change. This implies that our verification approach and analysis are different from traditional sensitivity analysis that uses proportional scaling which is not applicable to the CDM methodology. Furthermore, in such analysis, a change in the model parameters under consideration is, typically, due to a change in the subjective judgment of an expert whose opinion is used in model development as opposed to the changes due to diagnostic performance variations. We demonstrate the proposed verification approach by using synthetic formation flight data, and show that our CDM development method does not lead to a fault diagnosis model that is sensitive to small variation in its parameters.

Uncertainty and sensitivity decomposition of building energy models

As building energy modelling becomes more sophisticated, the amount of user input and the number of parameters used to define the models continue to grow. There are numerous sources of uncertainty in these parameters, especially when the modelling process is being performed before construction and commissioning. Past efforts to perform sensitivity and uncertainty analysis have focused on tens of parameters, while in this work, we increase the size of analysis by two orders of magnitude (by studying the influence of about 1000 parameters). We extend traditional sensitivity analysis in order to decompose the pathway as uncertainty flows through the dynamics, which identifies which internal or intermediate processes transmit the most uncertainty to the final output. We present these results as a method that is applicable to many different modelling tools, and demonstrate its applicability on an example EnergyPlus model.

Identifying alignment parameters affecting implanted patellofemoral mechanics

Complications of the patellofemoral (PF) joint remain a common cause for revision of total knee replacements. PF complications, such as patellar maltracking, subluxation, and implant failure, have been linked to femoral and patellar component alignment. In this study, a dynamic finite element model of an implanted PF joint was applied in conjunction with a probabilistic simulation to establish relationships between alignment parameters and PF kinematics, contact mechanics, and internal stresses. Both traditional sensitivity analysis and a coupled probabilistic and principal component analysis approach were applied to characterize relationships between implant alignment and resulting joint mechanics. Critical alignment parameters, and combinations of parameters, affecting PF mechanics were identified for three patellar designs (dome, modified dome, and anatomic). Femoral internal-external (I-E) alignment was identified as a critical alignment factor for all component designs, influencing medial-lateral contact force and anterior-posterior translation. The anatomic design was sensitive to patellar flexion-extension (F-E) alignment, while the dome, as expected, was less influenced by rotational alignment, and more by translational position. The modified dome was sensitive to a combination of superior-inferior, F-E, and I-E alignments. Understanding the relationships and design-specific dependencies between alignment parameters can aid preoperative planning, and help focus instrumentation design on those alignment parameters of primary concern.

A Practical Procedure for Customizable One-Way Sensitivity Analysis in Additive Value Models

Although a wide variety of sensitivity analysis methods for additive value models appear in the literature, the most commonly used and easily understood is one-way sensitivity analysis. We address concerns from practice with this traditional method and build the mathematical foundation for a new and robust one-way sensitivity analysis approach that enables the decision maker(s) to gain a greater ownership of the model. A traditional one-way sensitivity analysis varies one attribute weight while keeping all others proportionally constant; although useful, it also has limitations, particularly in group decision making. Our proposed methodology, customizable one-way sensitivity analysis (COSA), retains the simplicity of the traditional method but extends it to a more powerful form. COSA allows the decision maker(s) to tailor the sensitivity analysis to the desired decision context by assigning weight coefficients of elasticity to the attribute weights. These parameters let the decision maker choose how the adjusted weight change is distributed throughout the model. Furthermore, we provide a detailed discussion about performing sensitivity analysis on complex hierarchies with more than one tier and include an applied example to demonstrate COSA's usefulness. Overall, a consistent standard for the one-way sensitivity analysis of additive value models is provided for the decision analysis community.

Multi-objective design and tolerance allocation for single- and multi-level systems

In this work we develop a method to perform simultaneous design and tolerance allocation for engineering problems with multiple objectives. Most studies in existing literature focus on either optimal design with constant tolerances or the optimal tolerance allocation for a given design setup. Simultaneously performing both design and tolerance allocation with multiple objectives for hierarchical systems increases problem dimensions and raises additional computational challenges. A design framework is proposed to obtain optimal design alternatives and to rank their performances when variations are present. An optimality influence range is developed to aid design alternatives selections with an influence signal-to-noise ratio that indicates the accordance of objective variations to the Pareto set and an influence area that quantifies the variations of a design . An additional tolerance design scheme is implemented to ensure that design alternatives meet the target tolerance regions. The proposed method is also extended to decomposed multi-level systems by integrating traditional sensitivity analysis for uncertainty propagation with analytical target cascading. This work enables decision-makers to select their best design alternatives on the Pareto set using three measures with different purposes. Examples demonstrate the effectiveness of the method on both single- and multi-level systems.

A variance-based sensitivity index function for factor prioritization

Among the many uses for sensitivity analysis is factor prioritization—that is, the determination of which factor, once fixed to its true value, on average leads to the greatest reduction in the variance of an output. A key assumption is that a given factor can, through further research, be fixed to some point on its domain. In general, this is an optimistic assumption, which can lead to inappropriate resource allocation. This research develops an original method that apportions output variance as a function of the amount of variance reduction that can be achieved for a particular factor. This variance-based sensitivity index function provides a main effect sensitivity index for a given factor as a function of the amount of variance of that factor that can be reduced. An aggregate measure of which factors would on average cause the greatest reduction in output variance given future research is also defined and assumes the portion of a particular factors variance that can be reduced is a random variable. An average main effect sensitivity index is then calculated by taking the mean of the variance-based sensitivity index function. A key aspect of the method is that the analysis is performed directly on the samples that were generated during a global sensitivity analysis using rejection sampling. The method is demonstrated on the Ishigami function and an additive function, where the rankings for future research are shown to be different than those of a traditional global sensitivity analysis.

Advanced sensitivity analysis of the fuzzy assignment problem

This paper concentrates on sensitivity analysis of the fuzzy assignment problem (FAP). Since most real environments are uncertain, the FAP is more realistic than the assignment problem in application. Owing to the high degeneracy of the FAP, as that of the assignment problem, traditional sensitivity analysis, called Type I sensitivity analysis, which determines the range in which the current optimal basis remains optimal, is impractical. Hence, we attempt to perform other two types of advanced sensitivity analysis, called Type II and Type III sensitivity analysis, to overcome this problem. A labeling algorithm is then presented, where Type II sensitivity analysis is to determine the range of perturbation to keep the current optimal assignment remaining optimal, and Type III sensitivity analysis is to determine the range for which the rate of change of optimal value function remains unchanged. The procedure of the labeling algorithm is divided into two parts: one is when the unassigned cell is perturbed, and the other is when the assigned cell is perturbed. An example is presented to demonstrate that the labeling algorithm is a useful tool for determining the Type II and Type III sensitivity analysis of the FAP.

The utility of transient sensitivity for wildlife management and conservation: Bison as a case study

Developing effective management strategies is essential to conservation biology. Population models and sensitivity analyses on model parameters have provided a means to quantitatively compare different management strategies, allowing managers to objectively assess the resulting impacts. Inference from traditional sensitivity analyses (i.e., eigenvalue sensitivity methods) is only valid for a population at its stable age distribution, while more recent methods have relaxed this assumption and instead focused on transient population dynamics. However, very few case studies, especially in long-lived vertebrates where transient dynamics are potentially most relevant, have applied these transient sensitivity methods and compared them to eigenvalue sensitivity methods. We use bison ( Bison bison) at Badlands National Park as a case study to demonstrate the benefits of transient methods in a practical management scenario involving culling strategies. Using an age and stage-structured population model that incorporates culling decisions, we find that culling strategies over short time-scales (e.g., 1–5 years) are driven largely by the standing population distribution. However, over longer time-scales (e.g., 25 years), culling strategies are governed by reproductive output. In addition, after 25 years, the strategies predicted by transient methods qualitatively coincide with those predicted by traditional eigenvalue sensitivity. Thus, transient sensitivity analyses provide managers with information over multiple time-scales in contrast to the long time-scales associated with eigenvalue sensitivity analyses. This flexibility is ideal for adaptive management schemes and allows managers to balance short-term goals with long-term viability.

Private costs of delayed privatization of TOT Public Company Limited

PurposeThe purpose of this paper is to examine the private costs of delaying privatization of TOT Public Company Limited, a state‐owned enterprise telephony carrier.Design/methodology/approachTo address the private costs of a delay in terms of privatization, a counterfactual forecasting model has been constructed for the years 2007‐2012. The starting year of the evaluation is 2006, as this year was the deadline for privatization under the World Trade Organization (WTO) agreement. The forecasting model uses the traditional investment approach as well as scenario analyses.FindingsThe traditional investment approach and sensitivity analyses provide similar results. Both show that the benefit of delaying the privatization process could be higher than if the WTO‐agreed time process for TOT privatization were followed, as the growth of revenue sharing and from concessionaires would be higher than if TOT were privatized. There are no benefits in the long run, however, as the growth rate of revenue in the base case is lower than that in the privatization case, leading to a reduction in the present value of the base case over time: hence, the longer the delay in privatization, the slower the firm growth.Originality/valueThere are very few studies on Thailand's telecommunications sector. The results of this study are also useful to other developing countries.

Implications of ecological data constraints for integrated policy and livelihoods modelling: An example from East Kalimantan, Indonesia

Policy and human livelihoods modelling increasingly demands integrated research which requires ecological expertise. However, contributions from ecologists are often based on sparse data. Rather than discounting such data, in this paper we demonstrate how ecological modelling can effectively contribute to the development of policy recommendations in spite of data constraints. In a petrol subsidy analysis in East Kalimantan, Indonesia, we accounted for ecological data uncertainty by (a) assuming large parameter value ranges and (b) conducting a robustness test for policy recommendations. In addition to data scarcity, counter-intuitive results emerged emphasising the need for model validation. These counter-intuitive results indicated that decreasing petrol prices led to increased poverty. This informed a policy recommendation to prevent the reduction of petrol prices below IDR 5500 per litre. Using two key livelihood resources (fish and honey), we found that while a traditional sensitivity analysis suggested highly robust results, a robustness test indicated that policy recommendations would change if the incorrectness of parameter values approached 50%. The results show that ecological modelling can contribute effectively in spite of sparse data to guide policy, as well as identifying future research priorities.

Multiparametric linear programming: Support set and optimal partition invariancy

Traditional sensitivity and parametric analysis in linear optimization was based on preserving optimal basis. Interior point methods, however, do not converge to a basic solution (vertex) in general. Recently, there appeared new techniques in sensitivity analysis, which consist in preserving so called support set invariancy and optimal partition invariancy. This paper reflects the renascence of sensitivity and parametric analysis and extends single-parametric results to the case when there are multiple parameters in the objective function and in the right-hand side of equations. Multiparametric approach enables us to study more complex perturbation occurring in linear programs than the simpler sensitivity analysis does. We present a description of the set of admissible parameters under the mentioned invariances, and compare them with the classical optimal basis concept.

An advanced computational algorithm for systems analysis of tokamak power plants

A new computational algorithm for tokamak power plant system analysis is being developed for the ARIES project. The objective of this algorithm is to explore the most influential parameters in the physical, technological and economic trade space related to the developmental transition from experimental facilities to viable commercial power plants. This endeavor is being pursued as a new approach to tokamak systems studies, which examines an expansive, multi-dimensional trade space as opposed to traditional sensitivity analyses about a baseline design point. The new ARIES systems code consists of adaptable modules which are built from a custom-made software toolbox using object-oriented programming. The physics module captures the current tokamak physics knowledge database including modeling of the most-current proposed burning plasma experiment design (FIRE). The engineering model accurately reflects the intent and design detail of the power core elements including accurate and adjustable 3D tokamak geometry and complete modeling of all the power core and ancillary systems. Existing physics and engineering models reflect both near-term as well as advanced technology solutions that have higher performance potential. To fully assess the impact of the range of physics and engineering implementations, the plant cost accounts have been revised to reflect a more functional cost structure, supported by an updated set of costing algorithms for the direct, indirect, and financial cost accounts. All of these features have been validated against the existing ARIES-AT baseline case. The present results demonstrate visualization techniques that provide an insight into trade space assessment of attractive steady-state tokamaks for commercial use.

The diagnostic utility of multiple-level likelihood ratios

Clinicians are accustomed to interpreting diagnostic test scores in terms of sensitivity and specificity. Many clinicians also appreciate that sensitivity and specificity need to be interpreted in terms of local base rates (i.e., pretest probability). However, most neuropsychological tests contain a wide range of scores. Important diagnostic information may be sacrificed when valid test scores are reduced to the simple dichotomy of "positive" or "negative" diagnosis that underlies sensitivity and specificity analysis. The purpose of this study is to provide an introduction to multiple-level likelihood ratios, a method for preserving the information in a wider range of scores. These statistics are first described using a hypothetical example of dementia screening, then with patient data from an epilepsy surgery sample. Multiple-level likelihood ratios have several advantages over sensitivity and specificity analysis because they are applied across a wider range of diagnostic scores, and generalize to settings with different base rates. We suggest that the diagnostic validity of many psychological tests may be underestimated by relying solely on traditional dichotomous sensitivity and specificity analysis.

An Advanced Computational Approach to System Modeling of Tokamak Power Plants

A new computational model for fusion power plant system studies is being developed for the ARIES program. An operational design space has been created to explore the most influential parameters in the physical, technological and economic trade space related to the developmental transition from experimental facilities to viable commercial power plants. This allows examination of a multi-dimensional trade space as opposed to traditional sensitivity analyses about a baseline design point. The influence of multifunctional, highly dependent parameters can easily be visualized, which may highlight one or a few difficult-to-achieve parameters that would yield a highly acceptable design solution. The new ARIES systems code consists of adaptable physics, engineering and costing modules which capture the current tokamak knowledge database and reflect both near-term as well as advanced technology solutions that are higher risk but have higher performance potential. To fully assess the impact of the range of physics and engineering implementations, the plant cost accounts have been revised to reflect a more functional cost structure. All of these features have been validated against the highly respected ARIES-AT baseline. The present results demonstrate novel visualization techniques for trade space assessment of attractive tokamaks for commercial use.

Fast Statistical Analysis of Process Variation Effects Using Accurate PLL Behavioral Models

Using the behavioral model of a circuit to perform behavioral Monte Carlo simulation (BMCS) is a fast approach to estimate performance shift under process variation with detailed circuit responses. However, accurate Monte Carlo analysis results are difficult to obtain if the behavioral model is not accurate enough. Therefore, this paper proposes to use an efficient bottom-up approach to generate accurate process-variation-aware behavioral models of CPPLL circuits. Without blind regressions, only one input pattern in the extraction mode sufficiently obtains all required parameters in the behavioral model. A quasi-SA approach is also proposed to accurately reflect process variation effects. Considering generic circuit behaviors, the quasi-SA approach saves considerable simulation time for complicated curve fitting but still keeps estimation accuracy. The experimental results demonstrate that the proposed bottom-up modeling flow and quasi-SA equations provide similar accuracy as in the RSM approach, using less extraction cost as in the traditional sensitivity analysis approach.

Hierarchical demographic approaches for assessing invasion dynamics of non-indigenous species: An example using northern snakehead ( Channa argus)

Models of species’ demographic features are commonly used to understand population dynamics and inform management tactics. Hierarchical demographic models are ideal for the assessment of non-indigenous species because our knowledge of non-indigenous populations is usually limited, data on demographic traits often come from a species’ native range, these traits vary among populations, and traits are likely to vary considerably over time as species adapt to new environments. Hierarchical models readily incorporate this spatiotemporal variation in species’ demographic traits by representing demographic parameters as multi-level hierarchies. As is done for traditional non-hierarchical matrix models, sensitivity and elasticity analyses are used to evaluate the contributions of different life stages and parameters to estimates of population growth rate. We applied a hierarchical model to northern snakehead ( Channa argus), a fish currently invading the eastern United States. We used a Monte Carlo approach to simulate uncertainties in the sensitivity and elasticity analyses and to project future population persistence under selected management tactics. We gathered key biological information on northern snakehead natural mortality, maturity and recruitment in its native Asian environment. We compared the model performance with and without hierarchy of parameters. Our results suggest that ignoring the hierarchy of parameters in demographic models may result in poor estimates of population size and growth and may lead to erroneous management advice. In our case, the hierarchy used multi-level distributions to simulate the heterogeneity of demographic parameters across different locations or situations. The probability that the northern snakehead population will increase and harm the native fauna is considerable. Our elasticity and prognostic analyses showed that intensive control efforts immediately prior to spawning and/or juvenile-dispersal periods would be more effective (and probably require less effort) than year-round control efforts. Our study demonstrates the importance of considering the hierarchy of parameters in estimating population growth rate and evaluating different management strategies for non-indigenous invasive species.

Decision theory and real estate development: a note on uncertainty

Purpose – Real estate development appraisal is a quantification of future expectations. The appraisal model relies upon the valuer/developer having an understanding of the future in terms of the future marketability of the completed development and the future cost of development. In some cases the developer has some degree of control over the possible variation in the variables. However, other variables are totally dependent upon the vagaries of the market at the completion date. To try to address the risk of a different outcome to the one expected the developer will often carry out a sensitivity analysis on the development. This paper aims to look at the processes for so doing and the role of the decision maker in the analysis.Design/methodology/approach – Traditional sensitivity analysis has generally only looked at the best and worst scenarios and has focused on the anticipated or expected outcomes. This does not take into account uncertainty and the range of outcomes that can happen. A fuller analysis...

Structural robustness of biochemical network models-with application to the oscillatory metabolism of activated neutrophils

Sensitivity of biochemical network models to uncertainties in the model structure, with a focus on autonomously oscillating systems, is addressed. Structural robustness, as defined here, concerns the sensitivity of the model predictions with respect to changes in the specific interactions between the network components and encompass, for instance, uncertain kinetic models, neglected intermediate reaction steps and unmodelled transport phenomena. Traditional parametric sensitivity analysis does not address such structural uncertainties and should therefore be combined with analysis of structural robustness. Here a method for quantifying the structural robustness of models for systems displaying sustained oscillations is proposed. The method adopts concepts from robust control theory and is based on adding dynamic perturbations to the network of interacting biochemical components. In addition to providing a measure of the overall robustness, the method is able to identify specific network fragilities. The importance of considering structural robustness is demonstrated through an analysis of a recently proposed model of the oscillatory metabolism in activated neutrophils. The model displays small parametric sensitivities, but is shown to be highly unrobust to small perturbations in some of the network interactions. Identification of specific fragilities reveals that adding a small delay or diffusion term in one of the involved reactions, likely to exist in vivo, completely removes all oscillatory behaviour in the model.

Global sensitivity analysis of a 3D street canyon model—Part I: The development of high dimensional model representations

Traditional methods for global uncertainty and sensitivity analysis (SA) such as Monte Carlo methods based on full model simulations are often not suitable for application in environmental modelling due to the nonlinearity and computational complexity of the models. The high dimensional model representation (HDMR) method was developed to express the input–output relationships of a complex model with a high dimensional input space. HDMR provides a model replacement that can be easily employed within global SA. In this work an optimisation method is developed as an extension to the existing set of HDMR tools to improve the accuracy of the mapping process. An application from the field of urban flow and dispersion is chosen to demonstrate the effectiveness of the approach. Model replacements of a k– ε model simulating the flow field in an urban street canyon are constructed. The necessary computational effort for their construction is considerably lower than required by traditional global SA methods. First- and second-order sensitivity indices can then be calculated using the replacements without the need for additional full model runs. Comparison with a large sample of full model runs demonstrates that the output statistics of the full model are well represented by the model replacements. The proposed HDMR method therefore provides a powerful tool for general application to global SA of environmental models.

SENSITIVITY ANALYSIS OF OBJECTIVE FUNCTION COEFFICIENTS OF THE ASSIGNMENT PROBLEM

Information of sensitivity analysis, in a linear programming problem, is usually more important than the optimal solution itself. However, traditional sensitivity analysis, which perturbs exactly one coefficient and then determines the range preserving the optimality of the current optimal base, is impractical for the assignment problem. An optimal basic solution of the assignment problem is inherently degenerate, so it may be that the optimal base has changed but the optimal assignment remains unchanged. Furthermore, elements of a column (or row) in a cost matrix of assignment problem are usually closely related and change simultaneously, not uniquely. This paper focuses on two kinds of sensitivity analyses for the assignment problem. One is to determine the sensitivity range, over which the current optimal assignment or all the optimal assignments remain optimal, while perturbing the elements of one column (or row) in a cost matrix of the assignment problem simultaneously but dependently. The other is to perturb elements of one column (or row) in a cost matrix of the assignment problem simultaneously but independently. Numerical illustrations are presented to demonstrate that the approaches are useful in practice.