1、1Reliability Optimization Analysis based on Structure of Probabilistic Model（Science College and Institute of Intelligent Vision and Image Information， China Three Gorges University， Yichang， Hubei， 443002， P.R. China） Abstract. The implementation of a project contains a lot of uncertainty， by their
2、 nature， random uncertainty， uncertainty and human factors （such as peoples limitations of understanding things and design methods， the use of analog approximate calculation results） due to the uncertainty of the future to better improve the structural reliability， this paper， the probability model
3、reviewed structural reliability. Key words： structural reliability； probability model； optimization Introductions The reliability is the nature of the product during use without failure. Broadly speaking， the “reliability“ refers to the satisfaction of the users of the products or businesses trust.
4、This satisfaction level is from a subjective judgment， in order to make specific and quantitative judgment on product 2reliability， product reliability can be defined as within a predetermined time in a predetermined condition and， components （products） ， equipment or system stability complete funct
5、ion of the extent or nature. The reliability of the product actually used is called reliability. Operational reliability can be divided into the inherent reliability and reliability. The inherent reliability is the reliability of the manufacturer of the product design must be established， that in ac
6、cordance with the planning of the reliability， from the selection of raw materials and parts， design， manufacturing， testing， until the various stages of the product produced by the established reliability. Use reliability products， after the reliability of the packaging， transportation， storage， in
7、stallation， use， maintenance， and other factors. 1. Reliability factor Reliability factor contains the durability of reliability， maintainability， and reliability of the design of the three elements. A durability product without failure or long service life is durability. Serviceable product failure
8、 can quickly and easily through the maintenance or repair troubleshooting， is serviceable. Product maintainability great relationship with the structure of the product design reliability. Design 3reliability is key to determining product quality， due to the complexity of the human - machine system，
9、and there may be errors in the operation， and the operation of the use of such environmental factors affect the likelihood of errors still exist， so design must fully consider the product easy to use and easy to operate， which is the design reliability. In general， the more product is easy to operat
10、e， the smaller the possibility of failures and security problems caused by human error or other problems； From another perspective， the event of the failure or security issues， and to take the necessary measures and preventive measures are very important. 3.3 Durability Sufficient durability of the
11、structure in a predetermined working environment within a predetermined time period， the deterioration of the material properties will not cause the structure unacceptable failure probability. From engineering conceptually， sufficient durability means in the normal maintenance of the structure can u
12、nder the conditions of normal use to the specified design life span. Overall stability refers to occurs when and after the accidental events， building structure only local damage without progressive collapse. 4. Failure model of the structural system 4The failures of the structural system model are
13、classified as follows： the structure of the way （statically determinate and statically indeterminate） ， Component Failure nature （brittle， ductile） ， the series model， parallel model， and string - parallel model. 4.1 Series model If the structure of any one component failure， the entire structure fa
14、ilure， failure analysis of the type of structural system to the series model made from brittle component statically indeterminate structure series model. 4.2 Parallel model Member in one or more than one component failure， the remaining member or failure of ductile component， able to maintain the ov
15、erall structure of the function， all statically indeterminate structure failure analysis to parallel model. Figure1 Parallel model 4.3 Series - parallel model The statically indeterminate ductility member ultimate failure state of the structure is not limited to a structure， such structural system t
16、o series - parallel model. Figure2 Series - parallel model 5. Optimization of reliability optimization 55.1 The reliability index The reliability of the design is multiple objectives and multiple uncertainties. The objectives include safety， serviceability， durability and economy indicators. Without
17、 uncertainties include probabilistic uncertainty analysis probabilistic model of uncertainty， the uncertainty of the physical and mechanical model， the cross-section resistance of the basic assumptions of the formula and calculation model inaccuracies and human neglect and material properties， geome
18、tric parameters error. In engineering design， can not cover everything， consider these uncertainties and therefore should be made in the engineering design of the overall idea of optimization， performance reliability-based design methodology， and development of appropriate reliability design softwar
19、e， engineering design reliability theoretical development direction. 5.2 Reliability importance with the continuous advancement of the scientific and technological level the engineering structural complexity of the system in increasing the uncertainty of the performance along with more and more prom
20、inent. Therefore， people must carry out the analysis and decision-making under uncertainty premise. Traditional probabilistic reliability methods based on 6probability theory works dealing with uncertainty in the most commonly used method. With further research， people gradually discovered that in a
21、ddition to the randomness Also there is another important uncertainty - vagueness. Therefore highlights the importance of reliability optimization. 5.3 The model structure of reliability optimization Reliability-based structural optimization model is generally calculated by the model and the optimiz
22、ation of the reliability analysis model. Reliability model involving structural response calculation （structural analysis） and structural reliability estimates， including by the response of the structure parameters of the structural analysis model calculations to estimate the limit state function an
23、d the gradient of the basic parameters. Reliability estimates involving random parameters describe the failure level defined failure mode enumeration and reliability index estimation. Optimize the calculation model described including introducing optimization variables， objective function and constr
24、aints in the reliability index requirements， optimization of the iterative solution of the problem. Overall reliability optimization calculation model also includes the objective function and the optimization variable sensitivity calculation 7and constraint calculation. Reliability-based structural
25、optimization usually put the weight of the structure as the objective function. Reliability requirements associated with the possible failure modes of the structure as a constraint. With the total expected value of the structure of the initial cost， maintenance costs， and failure losses as the objec
26、tive function. At this point， you can design a more comprehensive consider the structure of the whole life of the expected total value and the optimal inspection and maintenance decisions. 5.4 Reliability optimization algorithm to optimize the Optimization algorithm， based on the sequential quadrati
27、c programming algorithm is generally considered to be the most effective and feasible method. Indirect method means to break the problem into several sub-structure （or subsystem）. Each subsystem has its own objective function and constraints. The various subsystems of the system from the low level c
28、ontrol action by the system is controlled at a high level， such methods involve the flow of information between the sub-structure and problems of coordination in order to achieve the overall goal. It should be noted that， regardless optimize a sub and sub interaction will lead to non-optimal solutio
29、ns. The well-known method of indirect model coordination method， 8target coordination method， linear decomposition method for structural reliability optimization designs some theoretical studies. Reliability optimization， optimal design， you need to repeatedly calculate the reliability of a componen
30、t or system. And effective nonlinear optimization algorithm requires an order； you must calculate the reliability or the reliability index sensitivity or gradient， the entire amount of its calculation. Therefore， the reliability calculation and optimization algorithm selection is very important. Eff
31、ective approximation method attention， through the expression of the structural response of these approximate methods， functions， structure and reliability of computing the approximate and simplified processing that the original complex optimization calculations can be accepted in the computation， a
32、nd can be used for the optimization of the complex structures. 6. The advantages of probabilistic model 6.1 The definition of probability models Giving user a query string， equals the presence of the string that contains the collection of all related documents. We see this collection as a desirable
33、outcome document set ideal result set is given； we can easily get the result document. So we can query processing as the treatment of the 9property of the desired result set of documents. The problem is that we do not know exactly what these attributes； we know there is an index expressed in terms o
34、f these attributes. Because these attributes are not visible during the query， which in the initial stage to estimate these properties. The estimate of this initial stage allows we return to the ideal result set for the first time to retrieve the document collection and generate an initial probabili
35、ty description. 6.2 Influencing factors of Probabilistic model Changes in the real world are being affected by a number of factors， including deterministic and random. From the background， purpose and means of modeling， the main factor is to determine the random factors can be ignored， or simply ran
36、dom factors can appear to the average effect， you will be able to establish a deterministic model. If random factors study must be taken into account， it should be the establishment of the stochastic model. 6.3 The advantages of probabilistic model The probabilistic model has the advantage that the
37、document can be a descending in accordance with their associated probabilities to calculate the Rank. His disadvantage： the need to guess at the beginning of the document is divided into 10two sets of relevant and irrelevant， in fact， this model does not consider the frequency of the index term in t
38、he document （all weights are binary） ， while the index terms are independent of each other. 7. Summaries The probability model is based on a given users query string and documents in the collection probability model to estimate the users query string probability associated with the document. The pro
39、bability model assumes that this probability is only determined by the query string and document. Further， the model assumes there is a collection of all the documents， that is， relative to the document subset of the result of the query string， this ideal collection represented by R， the documents i
40、n the collection is expected to query string. This feature can be introduced into the field of structural reliability optimization method to analyze and solve structural reliability. Acknowledgements Fund support： NSFC Grant 60972162 References 1.Chen jianqiao. Introduction to Mechanics of Composite MaterialsM. Science Press， 2006 2.An weiguang， Chai yinlin， Chen weidong. Random
