Opreations research seeks the determination of the best (optimum) course of action of a decision problem under the restriction of limited resources.
§@·~¬ã¨s¬O¦b¦³¸ê·½ªº¨î¤U´M¨D³Ì¨Î¤è®×ªº¤èªk¡C
Operations research is the use of quantitative models to analyze and predict the behavior of systems that are influenced by human decisions.
§@·~¬ã¨s¨Ï¥Î¶q¤Æªº¼Ò¦¡¨Ó¤ÀªR»P¹w´ú¨ü¤H¬°¨Mµ¦¼vÅTªº¨t²Îªº¦æ¬°¡C
OR quite often is associated with the use of mathematical techniques to model and analyze decision problems. (Science)
§@·~¬ã¨s³q±`¨Ï¥Î¼Æ¾Ç§Þ¥©¨Ó«Ø¥ß¨Ã¤ÀªR¨Mµ¦°ÝÃD¡C
Decision problems usually include important intangible factors that cannot be translated directly in terms of the mathematical model. (Art)
¨Mµ¦°ÝÃD³q±`¥]¬A«n«oÃø¥HÂà¤Æ¬°¼Æ¾Ç¼Ò¦¡ªº¥i¶q¤Æªº¦]¤l¡C
As a problem solving technique, OR must be viewed as both a science and an art. The science aspect lies in providing mathematical techniques and algorithms for solving appropriate decision problems. OR is an art because success in the phases that precede and succeed the solution of a mathematical model largely depends on the creativity and personal ability of the decision-making analysts.
°µ¬°¤@Ó´£¨Ñ¸Ñ¨M°ÝÃD¤èªkªº§Þ³N¡A§@·~¬ã¨s¥iµø¬°¬O¤@ªù¬JÃÀ³N¤]¬ì¾Çªº¾Ç°Ý¡A¦]¬°¦b¨D¸Ñ¬Y¨Ç¨Mµ¦°ÝÃD®É¯A¤Î¼Æ¾Ç§Þ¥©»Pµ¦²¤¡A©Ò¥H¬O¬ì¾Ç¡A¦ý¦b«Ø¥ß¼Ò¦¡¤§«e»P¨D¸Ñ¤§«áªº¸¨¹ê«h¯A¤Î¨Mµ¦¤ÀªRªÌªº³Ð³y¤O»PÓ¤H¯à¤O¡A©Ò¥H¤]¬OÃÀ³N¡C
Gathering of the data for model construction¸ê°Tªº¦¬¶° |
Art |
model construction, solution ¼Ò¦¡ªº«Ø¥ß¡B¨D¸Ñ |
Science |
Model validation ¼Ò¦¡ªºÅçÃÒ |
Art |
Implementation of the obtained solution µ²ªGªº¥I½Ñ¹ê¦æ |
Art |
The effect of human behavior has so influenced the decision problem that the solution obtained from the mathematical model is deemed impractical.
¤HÃþ¦æ¬°¹ï¨Mµ¦°ÝÃDªº¼vÅT¬O«Ü¤jªº¡A¦³®É¥Ñ¼Æ¾Ç¼Ò¦¡©Ò¨D¥Xªº¸Ñ¨M¤è®×³Qµø¬°¤£¥i¦æ¡C
Example: Elevator problem ¹q±è°ÝÃD: Solution from waiting line (queuing) model was found unsatisfactory. Problem solved by installing full-length mirrors.
Principal Components(¥Dn²Õ¦¨):objective ¥Ø¼Ð and variables ¨Mµ¦ÅܼÆ
An objective is the end result we desire to achieve by selecting a specific course of action for the system under study.
Maximize profit, flow
Minimize cost, time, distance or
better quality (intangible) are common objectives.
A decision problem arises because the decision maker is usually confronted with more than one course of action. Once an objective is defined, the optimum decision is selected as the course of action that best meets the objective. However, the 'quality' of the chosen decision depends on whether or not all the alternative courses of action are known to the decision maker.
The determination of possible courses of action represents a crucial aspect of the decision problem. It entails identifying the variables (governing parameters) of the system that are under the control of the decision-maker.
Two basic components are essential for constructing a model:
1. The objective of the system ¨t²Îªº¥Ø¼Ð
2. The constraints imposed on the system ¨t²Îªº¨î
Both the objective and constraints must be expressed in terms of the controllable variables (courses of action) of the system.
¥Ø¼Ð»P¨î¬Ò¶·¥H¨t²Î¤¤¥i±±¨îªºÅܼƨӪí¥Ü¡A¦¹¨ÇÅܼƥNªíµÛ¸Ñ¨M¤è®×ªº¶i¦æ¤è¦¡¡C
The analysis of the model should then yield the best course of action (in terms of the objective) that satisfies all the system's constraints.
¼Ò¦¡ªº¤ÀªR«h¬O¦b§ä¥Xº¡¨¬¨t²Îªº©Ò¦³¨î±ø¥ó¤Uªº³Ì¨Î¸Ñ¨M¤è®×¡C
1.3.1 Types of OR models§@·~¬ã¨s©Ò±´°Qªº¼Ò¦¡
l Symbolic or mathematical model (the most important type) ²Å¸¹¦¡ªº¼Ò¦¡©ÎºÙ¼Æ¾Ç¼Ò¦¡: It assumed that all the relevant variables, parameters and constraints as well as the objective are quantifiable. ¬O©Ò¦³¼Ò¦¡¤¤³Ì«nªº¤@ºØ¡A¦¹ºØ¼Ò¦¡°²³]©Ò¦³ªºÅܼơB°Ñ¼Æ¡B¨î»P¥Ø¼Ð¬Ò¥i¶q¤Æ¡C
l Simulation model ¼ÒÀÀ¼Ò¦¡: It 'imitate' the behavior of the system over a period of time. This is achieved by specifying a number of events that are points in time whose occurrence signifies that important information pertaining to the behavior of the system can be gathered (discrete event). Once such events are defined, it is necessary to pay attention to the system only when an event occurs. The information yielding ¡¥measures of performance¡¦ for the system is accumulated as statistical observations, which are updated as each event takes place. ¼ÒÀÀ¼Ò¦¡¥é·Ó¨t²Îªº¦æ¬°°õ¦æ¤@¬q®É¶¡¡A¨ä°µªk¬O©w¸q¤@¨Ç¨Æ¥ó (event)¨Ó¥Nªí¨t²Îªº¦æ¬°¡A¦¬¶°¤@¬q®É¶¡¤ºµo¥Í¦¹¨Ç¨Æ¥óªº¬ÛÃö¸ê°T(¨t²Îªºª¬ºA)§Y¥iÁA¸Ñ¨t²Îªº¦æ¬°¡C¦¹¨Ç¨Æ¥ó¨Ã¤£·|³sÄòµo¥Í©Ò¥H¤@¯ëºÙ¬°Â÷´²¨Æ¥ó(discrete event)¡C¥u¦³¦b¨t²Îµo¥Í¦¹¨Ç¨Æ¥ó®É¨äª¬ºA¤~·|¦³©Ò§ïÅÜ¡A¬ö¿ý¦¹¨Ç¥i±o¥X¨t²Î¦æ¬°«ü¼Ðªºª¬ºAÈ¡A³z¹L²Îp¤ÀªR¡A¥i§ó·s¨t²Î¦æ¬°ªº«ü¼ÐÈ¡C
l Heuristic model ±À²z¼Ò¦¡: Sometimes the mathematical formulation may be too complex to allow an exact solution or the required computation may be impractically long. In this case, heuristics can be used to develop good (approximate) solutions. Heuristics are search procedures that intelligently move from one solution point to another with the objective of improving the value of the model objective. When no further improvements can be achieved, the best-attained solution is the approximate solution to the model. ¦³®É¼Æ¾Ç¼Ò¦¡¤Ó¹L½ÆÂø©Î«ÜÃøpºâ¡A¦¹®É±À²z¼Ò¦¡¥i¥Î¨Ó¨D¥Xªñ¦ü¸Ñ¡C±À²z¬O¤@ºØ¦³´¼¼zªº·j´Mµ{§Ç¡A¥Ñ¤@ӸѴ¥t¤@Ó§ó¨Î¸ÑÁÚ¶i¡A©Ò¿×§ó¨Î«Y¥Ñ¼Ò¦¡¤¤©w¸qªº¥Ø¼Ð¨Ó¨M©w¡C·íµLªk¦A§ä¥X§ó¨Î¸Ñ®É¡A©Ò±oªº¸Ñ·¥¬°¨t²Îªºªñ¦ü¸Ñ¡C
1.3.2 Type of OR problems §@·~¬ã¨s©Ò±´°Qªº°ÝÃD
Sequencing ±Æ§Ç |
Allocation ¤À¬£ |
Routing (Networking) ¸ô®| (ºô¸ô) |
Replacement §ó·s |
Inventory ®w¦s |
Queuing ±Æ¶¤ |
Competitive Ävª§ |
Search ·j´M |
¦¹¨ÇÃþ«¬¤¤¤§¨C¤@Ãþ³£¤w¦³¼Æ¾Ç¼Ò¦¡¤§«Ø¥ß¡CµM¦Ó¡A·í½ÆÂø«×¼W¥[¡A¯Â²z½×Ãø¥H¨D¸Ñ®É¡A¡u¼ÒÀÀ¡v¡]Simulation¡^¥i¯à·|¬O°ß¤@¨D¥Xµª®×ªº¤èªk¡C
Sequencing ±Æ§Ç°ÝÃD¡G |
±´°Q¥H¤@Ó¤@©w¶¶§Ç©Î¦¸²Ä©ñ¸mª««~¥H«KªA°Èªº°ÝÃD¡C Ä´Y¡G¦b¤@Ó¤u§@¯¸ùØ¡AN Ó¤u§@¦b¤£¦Pªº¾÷¾¹ùػݤ£¦Pªº¤u§@®É¶¡¡A¨C¤@Ó¤u§@³£¥²»Ý¥H¦P¼Ëªº¦¸§Ç¦b M³¡¾÷¾¹¤W¥[¤u¦Ó¤£¯à¸õ¹L¡A«hÀ³¸Ó¦p¦ó¦w±Æ¦U¤u§@ªº¦¸§Ç¡A¥H¨Ï¦b©Ò¦³¾÷¾¹¤W¥þ³¡¤u§@ªº¥[¤uÁ`®É¶¡¬°³Ì¤Ö? |
Allocation ¤À°t°ÝÃD¡G |
¥H¬YºØµû¦ô®Ä²v¡]measure of effectiveness¡^ªºì«h¨Ó¤À°t¦UºØ§@·~ªº¸ê·½¡C Ä´Y¡G |
¤Hû |
Job 1 |
Job 2 |
Job 3 |
A |
2 |
6 |
3 |
B |
8 |
4 |
9 |
C |
5 |
7 |
8 |
Routing ¸ô®|°ÝÃD¡G |
±q°_ÂI¨ì²×ÂI¦³³\¦h¥i¦æªº¸ô®|¡An±q¤¤§ä¥X¤@±ø³Ì¨Îªº¸ô®|¡C |
Replacement §ó·s°ÝÃD¡G |
À³¦b¦ó®É§ó´«©Îºûרº¨Ç§Y±N³ø¼o©Î¤£³ô¥Îªº³]³Æ¡C |
Inventory ®w¦s°ÝÃD¡G |
¨M©w¬Y¨Ç²£«~©Î¹s²Õ¥ó¤§®w¦s¶q¦h¹è¡A»P¸É³f®É¾÷¡C |
Queuing ±Æ¶¤°ÝÃD¡G |
¤SºÙ¡uµ¥Ô½u¡v¡]Waiting line¡^°ÝÃD¡A¥ô¦ó¤@Ó¯A¤Îµ¥ÔªA°Èªº°ÝÃD§¡Äݤ§¡C ¿ì¤½¤j¼Ó»Ý´XÓ¹q±è?»È¦æÂd¥xµ¡¤fn¶}´XÓ? ¥[ªo¯¸n¦³´X½u? ¬Y¤½¥q¦³¤Q¥x¾÷¾¹¡A¦ý¦]¦~¤[»Ý¸g±`ºûסA»Ý¹µ¥Î´XÓ¾Þ§@û¡A´XÓײzû? |
Competitive Ävª§°ÝÃD¡G |
|
Search ·j´M°ÝÃD¡G |
|
·¥¤p¤Æ |
¦³Ãö¦¬¶°»P¤ÀªR¸ê®Æªº¦¨¥» |
¨Ã´î¤Ö |
¨Mµ¦¹Lµ{¥i¯àµo¥Íªº»~®t |
¶i¦Ó·¥¤p¤Æ |
¦³Ãö¨Mµ¦»~®t©Ò·l¥¢¤§¦¨¥» |
Table 1.1 Use of OR in current activities (1972 Turban)
¦UÃþ§@·~¬ã¨s¤èªk¦b¥Ø«eÀ³¥Î¤Wªº¨Ï¥ÎÀW²v
¤èªk |
±M®×¼Æ |
¨Ï¥ÎÀW²v |
²Îp¤ÀªR |
63 |
29 |
¼ÒÀÀ |
54 |
25 |
½u©Ê³W¹º |
41 |
19 |
¦s³f²z½× |
13 |
6 |
p¹ºµû®Ö³N |
13 |
6 |
°ÊºA³W¹º |
9 |
4 |
«D½u©Ê³W¹º |
7 |
3 |
±Æ¶¤²z½× |
2 |
1 |
±À²z³W¹º |
2 |
1 |
¨ä¥L |
13 |
6 |
Table 1.2 Relative use of OR techniques (1977 Ledbetter »PCox)
¦UÃþ§@·~¬ã¨s¤èªkªº¨Ï¥Îµ{«×
¤èªk |
±M®×¼Æ |
¨Ï¥Îµ{«× 1±q¥¼¥Î 5³Ì±`¥Î |
°jÂk¤ÀªR |
74 |
3.97 |
½u©Ê³W¹º |
78 |
3.36 |
¼ÒÀÀ |
70 |
3.31 |
ºô¸ô¼Ò¦¡ |
69 |
2.14 |
±Æ¶¤²z½× |
71 |
1.96 |
°ÊºA³W¹º |
69 |
1.62 |
¹ï§½²z½× |
67 |
1.61 |
Table 1.3 Quality of results by firms employing mathematical programming
¦U¤½¥q¨Ï¥Î¼Æ²z³W¹ºªºµ²ªG¤ÀªR (1976 Fabozzi »PValente)
|
½u©Ê³W¹º |
«D°ÊºA³W¹º |
°ÊºA³W¹º |
|||
µ²ªG |
¼Æ¥Ø |
% |
¼Æ¥Ø |
% |
¼Æ¥Ø |
% |
¨Î |
102 |
76 |
38 |
57 |
27 |
53 |
¥¢±Ñ |
21 |
16 |
19 |
28 |
15 |
29 |
®t |
6 |
3 |
6 |
9 |
3 |
6 |
¤£½T©w |
7 |
5 |
4 |
6 |
6 |
12 |
¦Xp |
133 |
100% |
67 |
100% |
51 |
100% |
Table 1.4 Educational background of OR personnel
§@·~¬ã¨s±q·~¤Hû¤§±Ð¨|I´º (1984 Hall)
¥D׾Ǭì |
¾Ç¤h |
ºÓ¤h |
³Õ¤h |
½Ñ¾Ç¦ì¦Xp |
§@·~¬ã¨s¤ÎºÞ²z¬ì¾Ç |
3 |
24 |
32 |
12 |
¼Æ¾Ç¤Î²Îp¾Ç |
26 |
16 |
21 |
22 |
¥ø·~ºÞ²z |
20 |
27 |
2 |
22 |
¤uµ{ |
34 |
17 |
29 |
28 |
¨ä¥L |
17 |
16 |
16 |
16 |
¦ûÁ`¼Æ¤§¦Ê¤À²v |
27 |
53 |
20 |
|
a. Problem Definition ©w¸q°ÝÃD
(1) description of the goal or the objective of the study,
©ú½T´yz¥Ø¼Ð
(2) identification of the decision alternatives of the system,
½T¥ß¨t²Îªº©Ò¦³¥i¯àªº¨Mµ¦¤è®×
(3) recognition of the limitations, restrictions, and requirements of the system.
ÁA¸Ñ¨t²Îªº©Ò¦³¨î»P»Ý¨D
b. Model Construction ¼Ò¦¡«Ø¥ß
c. Model Solution ¼Ò¦¡¨D¸Ñ
d. Model Validation ¼Ò¦¡ÅçÃÒ
e. Implementation of the final results ¼Ò¦¡¸¨¹ê