21233.rar_vehicle_modern control_其他书籍下载-pudn.com

<html xmlns="http://www.w3.org/1999/xhtml"> <head> <meta charset="utf-8"> <meta name="generator" content="pdf2htmlEX"> <meta http-equiv="X-UA-Compatible" content="IE=edge,chrome=1"> <link rel="stylesheet" href="https://static.pudn.com/base/css/base.min.css"> <link rel="stylesheet" href="https://static.pudn.com/base/css/fancy.min.css"> <link rel="stylesheet" href="https://static.pudn.com/prod/directory_preview_static/6265a6ef4c65f412590d1f22/raw.css"> <script src="https://static.pudn.com/base/js/compatibility.min.js"></script> <script src="https://static.pudn.com/base/js/pdf2htmlEX.min.js"></script> <script> try{ pdf2htmlEX.defaultViewer = new pdf2htmlEX.Viewer({}); }catch(e){} </script> <title></title> </head> <body> <div id="sidebar" style="display: none"> <div id="outline"> </div> </div> <div id="pf1" class="pf w0 h0" data-page-no="1"><div class="pc pc1 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/6265a6ef4c65f412590d1f22/bg1.jpg"><div class="t m0 x1 h2 y1 ff1 fs0 fc0 sc0 ls0 ws0">17 </div><div class="t m0 x2 h3 y2 ff1 fs1 fc0 sc0 ls1 ws1">Automatic Guided Vehicle Simulation </div><div class="t m0 x3 h3 y3 ff1 fs1 fc0 sc0 ls2 ws2">in MATLAB by Using Genetic Algorithm </div><div class="t m0 x4 h4 y4 ff2 fs2 fc0 sc0 ls3 ws3">Anibal Azevedo </div><div class="t m0 x5 h5 y5 ff3 fs3 fc0 sc0 ls4 ws4">State University of São Paulo </div><div class="t m0 x6 h5 y6 ff3 fs3 fc0 sc0 ls5 ws0">Brazil </div><div class="t m0 x7 h6 y7 ff1 fs3 fc0 sc0 ls6 ws5">1. Introduction </div><div class="t m0 x7 h7 y8 ff2 fs4 fc0 sc0 ls8 ws6">The prodigious advances in robotics in recent times made the use of robots more present in </div><div class="t m0 x7 h7 y9 ff2 fs4 fc0 sc0 ls8 ws8">modern society. One important advance that requires special attention is the development </div><div class="t m0 x7 h7 ya ff2 fs4 fc0 sc0 ls9 ws9">of an unmanned aerial vehicle (UAV), which allows an aircraft to fly without having a </div><div class="t m0 x7 h7 yb ff2 fs4 fc0 sc0 lsc ws0">human crew on board, although the UAVs still need to be controlled by a pilot or a </div><div class="t m0 x7 h7 yc ff2 fs4 fc0 sc0 lsd wsc">navigator. </div><div class="t m0 x7 h7 yd ff2 fs4 fc0 sc0 lse wsd">Today’s UAVs often combine remote control and computerized automation in a manner </div><div class="t m0 x7 h7 ye ff2 fs4 fc0 sc0 ls0 wsf">that built-in control and/or guidance systems perform deeds like speed control and flight-</div><div class="t m0 x7 h7 yf ff2 fs4 fc0 sc0 lsa ws10">path stabilization. In this sense, existing UAVs are not truly autonomous, mostly because </div><div class="t m0 x7 h7 y10 ff2 fs4 fc0 sc0 lsf ws12">air-vehicle autonomy field is a recently emerging field, and this could be a bottleneck for </div><div class="t m0 x7 h7 y11 ff2 fs4 fc0 sc0 ls11 ws14">future UAV development. </div><div class="t m0 x7 h7 y12 ff2 fs4 fc0 sc0 lsf ws15">It could be said that the ultimate goal in the autonomy technology development is to replace </div><div class="t m0 x7 h7 y13 ff2 fs4 fc0 sc0 ls12 ws16">human pilots by altering machines decisions in order to make decisions like humans do. For </div><div class="t m0 x7 h7 y14 ff2 fs4 fc0 sc0 ls13 ws17">this purpose, several tools related with artificial intelligence could be employed such as </div><div class="t m0 x7 h7 y15 ff2 fs4 fc0 sc0 ls10 ws18">expert systems, neural networks, machine learning and natural language processing </div><div class="t m0 x7 h7 y16 ff2 fs4 fc0 sc0 ls14 ws1a">(HAYKIN, 2009). Nowadays, the field of autonomy has mostly been following a bottom-up </div><div class="t m0 x7 h7 y17 ff2 fs4 fc0 sc0 ls10 ws1b">approach, such as hierarchical control systems (SHIM, 2000). </div><div class="t m0 x7 h7 y18 ff2 fs4 fc0 sc0 lsf ws1d">One interesting methodology from the hierarchical control systems approach is the </div><div class="t m0 x7 h7 y19 ff2 fs4 fc0 sc0 ls8 ws1f">subsumption architecture that decomposes complicated intelligent behavior into many </div><div class="t m0 x7 h7 y1a ff2 fs4 fc0 sc0 ls8 ws21">“simple” behavior modules, which are organized into layers. Each layer implements a </div><div class="t m0 x7 h7 y1b ff2 fs4 fc0 sc0 ls11 ws23">particular goal and higher layers are increasingly more abstract. The decisions are not taken </div><div class="t m0 x7 h7 y1c ff2 fs4 fc0 sc0 ls17 ws25">by a superior layer, but by listening to the information that are triggered by sensory inputs </div><div class="t m0 x7 h8 y1d ff2 fs4 fc0 sc0 ls18 ws27">(lowest layer). This methodology allows the use of reinforcement learning to acquire </div><div class="t m0 x7 h7 y1e ff2 fs4 fc0 sc0 ls11 ws14">behavior with the information that comes with experience. </div><div class="t m0 x7 h8 y1f ff2 fs4 fc0 sc0 ls19 ws2a">Inspired by old behaviorist psychology, reinforcement learning (RL) concerned with how </div><div class="t m0 x7 h7 y20 ff2 fs4 fc0 sc0 ls19 ws0">an agent ought to take actions in an environment, so as to maximize some notion of </div><div class="t m0 x7 h7 y21 ff2 fs4 fc0 sc0 ls1e ws0">cumulative reward. Reinforcement learning differs from standard supervised learning in </div><div class="t m0 x7 h7 y22 ff2 fs4 fc0 sc0 ls9 ws30">that correct input/output pairs which are never presented (HAYKIN, 2009). Furthermore, </div><div class="t m0 x7 h7 y23 ff2 fs4 fc0 sc0 ls17 ws32">there is a focus on an on-line performance, which involves finding a balance between </div><div class="t m0 x7 h7 y24 ff2 fs4 fc0 sc0 lsf ws34">exploration (of uncharted territory) and exploitation (of current knowledge). The </div><div class="t m0 x7 h7 y25 ff2 fs4 fc0 sc0 ls20 ws36">reinforcement learning has been applied successfully to various problems, including robot </div><div class="t m0 x7 h7 y26 ff2 fs4 fc0 sc0 ls14 ws38">control, elevator scheduling, telecommunications, backgammon and chess (SHIM, 2000). </div><div class="t m0 x8 h9 y27 ff5 fs5 fc1 sc0 ls18 ws0">www.intechopen.com</div></div><div class="pi" data-data='{"ctm":[1.991701,0.000000,0.000000,1.991701,0.000000,0.000000]}'></div></div> </body> </html>

<div id="pf2" class="pf w0 h0" data-page-no="2"><div class="pc pc2 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/6265a6ef4c65f412590d1f22/bg2.jpg"><div class="t m0 x9 ha y28 ff6 fs6 fc0 sc0 ls18 ws0"> </div><div class="t m0 xa ha y29 ff6 fs6 fc0 sc0 ls21 ws3a">MATLAB for Engineers – Applications in Control, Electrical Engineering, IT and Robotics </div><div class="t m0 x7 h7 y2a ff6 fs4 fc0 sc0 ls22 ws0">410 </div><div class="t m0 x7 h8 y2b ff4 fs4 fc0 sc0 lsa ws3c">Genetic algorithms (GAs) are developed in order to emulate the process of genetic </div><div class="t m0 x7 h7 y2c ff2 fs4 fc0 sc0 lsd ws3e">evolution found in nature, and then perform artificial evolution. They were developed by </div><div class="t m0 x7 h7 y2d ff2 fs4 fc0 sc0 ls10 ws40">John Holland [11] in the early 70s, and have been successfully applied to numerous large </div><div class="t m0 x7 h7 y2e ff2 fs4 fc0 sc0 ls18 ws41">and complex search space problems ever since (MICHALEWICZ, 1996). </div><div class="t m0 x7 h7 y2f ff2 fs4 fc0 sc0 ls23 ws43">In nature, organisms have certain characteristics that affect their ability to survive and </div><div class="t m0 x7 h7 y30 ff2 fs4 fc0 sc0 ls24 ws44">reproduce. These characteristics are contained in their genes. Natural selection ensures </div><div class="t m0 x7 h7 y31 ff2 fs4 fc0 sc0 ls26 ws46">that genes from a strong individual are presented in greater numbers in the next </div><div class="t m0 x7 h7 y32 ff2 fs4 fc0 sc0 ls28 ws48">generation, rather than those from a weak individual. Over a number of generations, the </div><div class="t m0 x7 h7 y33 ff2 fs4 fc0 sc0 ls25 ws4a">fittest individuals, in the environment in which they live, have the highest probability of </div><div class="t m0 x7 h7 y34 ff2 fs4 fc0 sc0 ls2a ws4c">survival and tend to increase in number; while the less fit individuals tend to die out. This </div><div class="t m0 x7 h7 y35 ff2 fs4 fc0 sc0 ls2c ws42">is the Darwin’s principle of the survival-of-the-fittest and constitutes the basic idea behind </div><div class="t m0 x7 h7 y36 ff2 fs4 fc0 sc0 ls2d ws4f">the GAs. In order to perform computational tests on how the reinforcement learning </div><div class="t m0 x7 h7 y37 ff2 fs4 fc0 sc0 ls2e ws50">could cope with genetic algorithms to provide good rules for the navigation of an </div><div class="t m0 x7 h7 y38 ff2 fs4 fc0 sc0 ls2f ws51">automatic vehicle (STAFYLOPATIS, 1998), a program that emulates a navigation </div><div class="t m0 x7 h7 y39 ff2 fs4 fc0 sc0 ls30 ws52">environment was developed in a Matlab. This Chapter will describe how the ideas </div><div class="t m0 x7 h7 y3a ff2 fs4 fc0 sc0 ls30 ws54">developed by (STAFYLOPATIS, 1998) could be employed to study the integration of the </div><div class="t m0 x7 h7 y3b ff2 fs4 fc0 sc0 ls30 ws55">GAs and RL to produce rules for automatic guided vehicles searching for a better </div><div class="t m0 x7 h7 y3c ff2 fs4 fc0 sc0 ls30 ws56">performance. The main contributions of this Chapter are: the vehicle, its sensors, and also </div><div class="t m0 x7 h7 y3d ff2 fs4 fc0 sc0 ls32 ws57">the environment for training are different from the ones presented in (STAFYLOPATIS, </div><div class="t m0 x7 h7 y3e ff2 fs4 fc0 sc0 ls2f ws59">1998); a new equation for the reinforcement learning was proposed; the influence of the </div><div class="t m0 x7 h7 y3f ff2 fs4 fc0 sc0 ls33 ws5b">parameters that control the production of automatic generation of rules for vehicle control </div><div class="t m0 x7 h7 y40 ff2 fs4 fc0 sc0 ls2d ws30">navigation are also tested. Sections 2 and 3 describe how the navigation decision rules </div><div class="t m0 x7 h7 y41 ff2 fs4 fc0 sc0 ls30 ws5d">could be encoded in a vector. Sections 4 and 5 show how the reinforcement learning and </div><div class="t m0 x7 h7 y42 ff2 fs4 fc0 sc0 ls35 ws5e">genetic algorithm uses the encoding of Section 2 to cope with the production of vehicle </div><div class="t m0 x7 h7 y43 ff2 fs4 fc0 sc0 ls2e ws60">control navigation rules. In Section 6, some results are presented and finally, in Section 7, </div><div class="t m0 x7 h7 y44 ff2 fs4 fc0 sc0 ls39 ws63">some conclusions and future works are given. All the proposed approach had been </div><div class="t m0 x7 h7 y45 ff2 fs4 fc0 sc0 ls33 ws64">implemented in a Matlab. </div><div class="t m0 x7 h6 y46 ff1 fs3 fc0 sc0 ls18 ws65">2. Navigation problem representation </div><div class="t m0 x7 h7 y47 ff2 fs4 fc0 sc0 ls3a ws66">The navigation problem could be defined as how a vehicle could be guided through an </div><div class="t m0 x7 h7 y48 ff2 fs4 fc0 sc0 ls11 ws68">ambient with many obstacles and barriers using the information available from the </div><div class="t m0 x7 h7 y49 ff2 fs4 fc0 sc0 lse ws6a">information given by the vehicle radar, as it can be seen in Figure 1. </div><div class="t m0 x7 h7 y4a ff2 fs4 fc0 sc0 ls17 ws6c">Figure 1 also details how the vehicle radar works. The radar has 9 sectors in order to better </div><div class="t m0 x7 h7 y4b ff2 fs4 fc0 sc0 ls17 ws6d">acquire the information of how near is a vehicle to an object. The information of existence of </div><div class="t m0 x7 h8 y4c ff2 fs4 fc0 sc0 ls3a ws6e">an obstacle for each section is stored in vector v of nine positions using the following rule: if </div><div class="t m0 x7 h7 y4d ff2 fs4 fc0 sc0 ls8 ws70">one object is in one section, then value 1 is attributed. If not, then value 0 is attributed. </div><div class="t m0 x7 h7 y4e ff2 fs4 fc0 sc0 ls17 ws71">Figure 2 details a situation where there are obstacles in sector 1 and 2 and the correspondent </div><div class="t m0 x7 h7 y4f ff2 fs4 fc0 sc0 ls8 ws72">representation by a vector. Once the objects had been detected by the radar, it is necessary to </div><div class="t m0 x7 h7 y50 ff2 fs4 fc0 sc0 ls18 ws41">implement an appropriate action which could be one of the following three: turn 15</div><div class="t m0 xb hb y51 ff2 fs7 fc0 sc0 ls18 ws74">o</div><div class="t m0 xc h7 y52 ff2 fs4 fc0 sc0 ls3c ws75"> degrees </div><div class="t m0 x7 h7 y53 ff2 fs4 fc0 sc0 ls17 ws76">to the left, keep the trajectory or turn 15</div><div class="t m0 xd hb y54 ff2 fs7 fc0 sc0 ls18 ws76">o</div><div class="t m0 xe h7 y53 ff2 fs4 fc0 sc0 ls1c ws77"> degrees to the right. For the sake of clarity, but </div><div class="t m0 x7 h7 y55 ff2 fs4 fc0 sc0 lsa ws78">without loss of generality, velocity v</div><div class="t m0 xf hb y56 ff2 fs7 fc0 sc0 ls18 ws78">c</div><div class="t m0 x10 h7 y57 ff2 fs4 fc0 sc0 ls8 ws6"> of the vehicle was assumed to be constant. With these </div><div class="t m0 x7 h8 y58 ff2 fs4 fc0 sc0 ls3a ws79">three possible decisions, a new concept could be created, which is a rule. A rule is a vector </div><div class="t m0 x7 h7 y59 ff2 fs4 fc0 sc0 ls14 ws7b">that combines the one that describes the situation for the vehicle in terms of obstacles, like </div><div class="t m0 x7 h7 y5a ff2 fs4 fc0 sc0 lsa ws7d">the one presented in Figure 2, and a new component that decides the vehicle action in turn </div><div class="t m0 x7 h7 y5b ff2 fs4 fc0 sc0 ls20 ws80">to avoid a set of obstacles: 0 – turn left, 1 – keep the trajectory, 2 – turn right. An example of </div><div class="t m0 x7 h7 y5c ff2 fs4 fc0 sc0 ls3a ws73">a rule for the situation showed in Figure 2 is described in Figure 3. </div><div class="t m0 x8 h9 y27 ff5 fs5 fc1 sc0 ls18 ws0">www.intechopen.com</div></div><div class="pi" data-data='{"ctm":[1.991701,0.000000,0.000000,1.991701,0.000000,0.000000]}'></div></div>

<div id="pf3" class="pf w0 h0" data-page-no="3"><div class="pc pc3 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/6265a6ef4c65f412590d1f22/bg3.jpg"><div class="t m0 x7 ha y28 ff6 fs6 fc0 sc0 ls18 ws0"> </div><div class="t m0 x7 ha y29 ff6 fs6 fc0 sc0 ls3d ws82">Automatic Guided Vehicle Simulation in MATLAB by Using Genetic Algorithm </div><div class="t m0 x11 h7 y2a ff6 fs4 fc0 sc0 ls22 ws0">411 </div><div class="t m0 x12 hc y5d ff2 fs8 fc0 sc0 ls18 ws0"> </div><div class="t m0 x7 h7 y5e ff2 fs4 fc0 sc0 ls3a ws73">Fig. 1. Description of the elements of the developed environment in Matlab. </div><div class="t m0 x7 h4 y5f ff2 fs2 fc0 sc0 ls18 ws0"> </div><div class="t m0 x13 h7 y60 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x14 h7 y61 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 xe h7 y62 ff2 fs4 fc0 sc0 ls8 ws0">Situation</div><div class="t m0 x14 h7 y63 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x15 h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x15 h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">1</div><div class="t m0 x16 h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x17 h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">1</div><div class="t m0 x5 h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x18 h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">0</div><div class="t m0 x19 h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1a h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">0</div><div class="t m0 x1b h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1c h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">0</div><div class="t m0 x1d h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1e h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">0</div><div class="t m0 x1f h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x20 h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">0 </div><div class="t m0 x21 h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x22 h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">0 </div><div class="t m0 x23 h7 y64 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x24 h7 y65 ff2 fs4 fc0 sc0 ls18 ws0">0 </div><div class="t m0 x14 h7 y66 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x25 h7 y67 ff2 fs4 fc0 sc0 lsa ws0">Vector </div><div class="t m0 x26 h7 y68 ff2 fs4 fc0 sc0 ls3f ws0">Index </div><div class="t m0 x14 h7 y69 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x27 h7 y6a ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x27 h7 y6b ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x15 h7 y6c ff2 fs4 fc0 sc0 ls18 ws0">0</div><div class="t m0 x16 h7 y6d ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x16 h7 y6e ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x17 h7 y6f ff2 fs4 fc0 sc0 ls18 ws0">1</div><div class="t m0 x5 h7 y70 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x5 h7 y71 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x18 h7 y72 ff2 fs4 fc0 sc0 ls18 ws0">2</div><div class="t m0 x19 h7 y73 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x19 h7 y74 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1a h7 y75 ff2 fs4 fc0 sc0 ls18 ws0">3</div><div class="t m0 x1b h7 y76 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1b h7 y77 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1c h7 y78 ff2 fs4 fc0 sc0 ls18 ws0">4</div><div class="t m0 x1d h7 y79 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1d h7 y7a ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1e h7 y7b ff2 fs4 fc0 sc0 ls18 ws0">5</div><div class="t m0 x1f h7 y7c ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x1f h7 y7d ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x20 h7 y7e ff2 fs4 fc0 sc0 ls18 ws0">6 </div><div class="t m0 x21 h7 y7f ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x21 h7 y80 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x22 h7 y81 ff2 fs4 fc0 sc0 ls18 ws0">7 </div><div class="t m0 x23 h7 y82 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x23 h7 y83 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x24 h7 y84 ff2 fs4 fc0 sc0 ls18 ws0">8 </div><div class="t m0 x28 h7 y85 ff2 fs4 fc0 sc0 ls40 ws84">Sector 1234567 8 9 </div><div class="t m0 x29 hd y86 ff2 fs9 fc0 sc0 ls18 ws0"> </div><div class="t m0 x7 h7 y87 ff2 fs4 fc0 sc0 ls11 ws14">Fig. 2. The correspondence between the obstacles detected by a radar vehicle and a vector </div><div class="t m0 x7 h7 y88 ff2 fs4 fc0 sc0 ls41 ws85">with binary information. </div><div class="t m0 x7 h4 y89 ff2 fs2 fc0 sc0 ls18 ws0"> </div><div class="t m0 x2a he y8a ff7 fs6 fc0 sc0 ls18 ws0"> </div><div class="t m0 x2a he y8b ff7 fs6 fc0 sc0 ls18 ws0"> </div><div class="t m0 x2b hf y8c ff8 fsa fc0 sc0 ls18 ws0"> </div><div class="t m0 x2c hf y8d ff8 fsa fc0 sc0 ls18 ws0"> </div><div class="t m0 x7 h7 y8e ff2 fs4 fc0 sc0 ls11 ws14">Fig. 3. The correspondence between a rule and decision taken by the vehicle. </div><div class="t m0 x8 h9 y27 ff5 fs5 fc1 sc0 ls18 ws0">www.intechopen.com</div></div><div class="pi" data-data='{"ctm":[1.991701,0.000000,0.000000,1.991701,0.000000,0.000000]}'></div></div>

<div id="pf4" class="pf w0 h0" data-page-no="4"><div class="pc pc4 w0 h0"><img class="bi x0 y0 w1 h1" alt="" src="https://static.pudn.com/prod/directory_preview_static/6265a6ef4c65f412590d1f22/bg4.jpg"><div class="t m0 x9 ha y28 ff6 fs6 fc0 sc0 ls18 ws0"> </div><div class="t m0 xa ha y29 ff6 fs6 fc0 sc0 ls21 ws3a">MATLAB for Engineers – Applications in Control, Electrical Engineering, IT and Robotics </div><div class="t m0 x7 h7 y2a ff6 fs4 fc0 sc0 ls22 ws0">412 </div><div class="t m0 x7 h6 y8f ff1 fs3 fc0 sc0 ls4 ws86">3. Navigation control </div><div class="t m0 x7 h7 y90 ff2 fs4 fc0 sc0 lsa ws78">The previous binary representation scheme has a main drawback since before every vehicle </div><div class="t m0 x7 h10 y91 ff2 fs4 fc0 sc0 ls42 ws87">movement decision demanded to store 3×2</div><div class="t m0 x2d hb y92 ff2 fs7 fc0 sc0 ls18 ws87">9</div><div class="t m0 x2e h7 y91 ff2 fs4 fc0 sc0 ls14 ws88"> = 1536 vectors (rules) to precisely describe </div><div class="t m0 x7 h7 y93 ff2 fs4 fc0 sc0 ls20 ws89">which action should be performed for each scenario detected by the radar information. This </div><div class="t m0 x7 h7 y94 ff2 fs4 fc0 sc0 lsa ws8a">implies that before every step performed by the vehicle, a control action should compare, in </div><div class="t m0 x7 h7 y95 ff2 fs4 fc0 sc0 ls18 ws8b">the worst case, 1536 vectors in order to find the proper decision to be taken. This scheme is </div><div class="t m0 x7 h7 y96 ff2 fs4 fc0 sc0 ls15 ws8c">illustrated in Fig. 4. </div><div class="t m0 x7 h7 y97 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x2f h7 y98 ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x7 h7 y99 ff2 fs4 fc0 sc0 ls1a ws8d">Fig. 4. Decision scheme followed by the automatic guided vehicle. </div><div class="t m0 x7 h7 y9a ff2 fs4 fc0 sc0 ls8 ws8e">This computational work could be avoided by introducing a new symbol 2, for the radar </div><div class="t m0 x7 h7 y9b ff2 fs4 fc0 sc0 ls11 ws8f">information section, which means that “it could have or not an obstacle at this sector”. The </div><div class="t m0 x7 h7 y9c ff2 fs4 fc0 sc0 ls8 ws54">advantage of the definition of this new symbol is that it will help to reduce the number of </div><div class="t m0 x7 h7 y9d ff2 fs4 fc0 sc0 ls14 ws91">necessary information to be kept by the vehicle control. The disadvantage of using this new </div><div class="t m0 x7 h7 y9e ff2 fs4 fc0 sc0 ls43 ws92">symbol is that it could group different situations where the decisions should not be the </div><div class="t m0 x7 h7 y9f ff2 fs4 fc0 sc0 ls44 ws94">same. Fig. 5 gives an example of this situation. </div><div class="t m0 x7 h7 ya0 ff2 fs4 fc0 sc0 ls17 ws6d">The example showed in Fig. 5 emphasizes the importance to construct the rules in a manner </div><div class="t m0 x7 h7 ya1 ff2 fs4 fc0 sc0 lsa ws78">that the parts of the rule which do not affect the decision should be numbered as 2, and the </div><div class="t m0 x7 h7 ya2 ff2 fs4 fc0 sc0 lsf ws95">other parts that have a great influence in the final behavior of the vehicle should be </div><div class="t m0 x7 h7 ya3 ff2 fs4 fc0 sc0 ls8 ws3f">numbered as most specific as possible or, in other words, with the numbers 0 or 1. Fig. 6 </div><div class="t m0 x7 h7 ya4 ff2 fs4 fc0 sc0 ls13 ws98">illustrates the application of this concept. </div><div class="t m0 x7 h7 ya5 ff2 fs4 fc0 sc0 ls46 ws99">Fig. 6 also gives a guideline procedure for another problem that could emerge, which is the </div><div class="t m0 x7 h7 ya6 ff2 fs4 fc0 sc0 ls9 ws9a">appearance of more than one rule that matches the current radar information when symbol </div><div class="t m0 x7 h7 ya7 ff2 fs4 fc0 sc0 ls3a ws9c">2 is used. One criteria that will be adopted is to select the rules which match the </div><div class="t m0 x7 h7 ya8 ff2 fs4 fc0 sc0 lsf ws9e">environment situation, but with as much specific information as possible. For this purpose </div><div class="t m0 x7 h7 ya9 ff2 fs4 fc0 sc0 ls16 ws9f">Eq. (1) will be adopted. </div><div class="t m0 x30 h7 yaa ff2 fs4 fc0 sc0 ls18 ws0"> </div><div class="t m0 x31 h11 yab ff2 fsb fc0 sc0 ls47 ws0">()</div><div class="t m0 x25 h12 yac ff3 fsc fc0 sc0 ls18 ws0">i</div><div class="t m0 x32 h12 yad ff3 fsc fc0 sc0 ls18 ws0">i</div><div class="t m0 x33 h13 yab ff3 fsb fc0 sc0 ls48 ws0">nk</div><div class="t m0 x34 h13 yae ff3 fsb fc0 sc0 ls18 ws0">S</div><div class="t m0 x13 h13 yaf ff3 fsb fc0 sc0 ls18 ws0">n</div><div class="t m0 x35 h14 yb0 ff9 fsb fc0 sc0 ls18 ws0">−</div><div class="t m0 x36 h14 yb1 ff9 fsb fc0 sc0 ls18 ws0">=</div><div class="t m0 x37 h7 yb2 ff2 fs4 fc0 sc0 ls49 ws0"> </div><div class="t m0 x38 h15 yb3 ff8 fsd fc0 sc0 ls18 ws0">1,,iR= </div><div class="t m0 x16 h7 yb2 ff2 fs4 fc0 sc0 lsa wsa0"> (1) </div><div class="t m0 x8 h9 y27 ff5 fs5 fc1 sc0 ls18 ws0">www.intechopen.com</div></div><div class="pi" data-data='{"ctm":[1.991701,0.000000,0.000000,1.991701,0.000000,0.000000]}'></div></div>