Reliability Engineering and Risk Analysis

P&R software is distributed with the book. This, in turn, requires MATLAB software, including the signal processing and control toolboxes, from Mathworks. Updates to the P&R software will be made available on the publisher’s Web site as needed. In general, the software works as indicated in Appendix A of the book. Some hints on using additional MATLAB capabilities are given in Appendix B. Various examples throughout the text include instructions on the use of particular functions. Some of the computer exercises call for the writing of MATLAB programs. The functions include the capability to generate random variables or analyze data for any speciŽ ed distributional form from the nine choices. The reader should be aware that some menus ask for variances and others for standard deviations. Menus sometimes ask for a percentage, but a proportion is the intended meaning. Since no tables of statistical distributions are provided in the text, students will need access to P&R or another source of statistical tables. This book is terse, like others in this genre. For example, the following are all introduced in fewer than eight lines of text and formulas on page 230—the sample median and sample range, a use of the median as an estimator of the mean, and use of a range-based estimator of the variance, together with the order statistics necessary to their computation. The statistics chapter is not for the fainthearted. If an instructor wishes, following a suggestion in the Preface, to expand the coverage of this material, some supplemental handouts will be needed. Interval estimates appear rather abruptly; they are never used in the discussion of linear regression. Particularly in the later chapters, the emphasis is on random processes for signal recognition. Signal-to-noise ratios receive prominent coverage. The text is not always in standard American English. Most of the grammatical lapses are failures in agreement of number between subject and predicate. I had little difŽ culty in supplying the appropriate meaning. There are some minor typographical errors in problems and examples. Problem 2.13 should specify sampling without replacement. In Problem 8.1, the output should be y4t5. In Example 4.23, parts (d) and (e) are interchanged. An incorrect probability calculated in part (c) of Example 2.24 is among the more serious errors; the correct approach is obvious if one considers the probability of the complementary event. A one-semester lesson plan is provided, with some suggestions on how to trim the 46 suggested lectures to Ž t into a 14-week term with three 50-minute classes per week. There are also suggestions for expanded coverage, such as were noted for the statistics chapter. There is good coverage of selected distributions and methods for generating pseudorandom numbers. These distributions are the normal, lognormal, chisquared uniform, exponential, Weibull, and Rayleigh for continuous variables and the binomial and Poisson for integers. The examples make use of these distributions in topics familiar to engineers, such as reliability and quality control The comparison of textbooks for a one-semester sequence for undergraduate electrical engineering students might be helpful. In that sense, the reviewer would exclude Balakrishnan’s (1995) book, reviewed by Sengupta (1997), since it is clearly for a second semester in a two-semester sequence on variability in processes. Ziemer’s (1997) book is closer to a more traditional statistics course, but it still omits many of the topics in Hogg’s (1994) course outline. A full review was provided by Samaranakaye and Patel (1998). Ziemer’s book has much less on signal processing than does Li’s. Like Li’s book, Ziemer’s is quite terse, but it has fewer applied examples than Li’s. Leon-Garcia’s (1994) book was a revision of the Ž rst edition (1989), which was reviewed by Leigh (1991). This could be used in a variety of one-semester or two-semester courses. A two-semester sequence would include not only probability, a little on statistics, and an overview of random processes but also Markov chains and queuing theory, which are important for computer engineering majors. A one-semester syllabus, if it were to include probability, statistics, and random processes, would need to be very selective in Chapters 3, 5, 6, and 7. This too is a tersely written book. On the whole, Li’s is a solid book to consider for a limited audience. The omission of queuing theory, if there were time to cover the topic, could weigh against recommending it for computer engineering majors, but it should be a good introduction to variation in random processes for most electrical or computer engineering students.