67:
The use of preferences may also increase the length of a plan in order to satisfy more preferences. For example, when planning a journey from home to school, the user may prefer to buy a cup of coffee along the way. The planning software could now plan to visit the coffee shop first and then continue
59:
Preferences can be regarded as soft constraints on a plan. The quality of a plan increases when more preferences are satisfied but it may not be possible to satisfy all preferences in one plan. This differs from hard constraints which must be satisfied in all plans produced by the planning software.
35:
as possible. In many problem domains, a task can be accomplished by various sequences of actions (also known as plans). These plans can vary in quality: there can be many ways to solve a problem, but preferred generally are ways more, e.g., cost-effective, quick, and safe.
137:
In addition to determining whether a preference is satisfied, we also need to compute the quality of a plan based on how many preferences are satisfied. For this purpose, PDDL 3.0 includes an expression called
142:
which is equal to "the number of distinct preferences with the given name that are not satisfied in the plan". For a plan, a value can now be computed using a metric function, which is specified with
64:
while the soft constraints (or preferences) are separately specified by the user. This allows the same domain knowledge to be reused for various users who may have different preferences.
282:
39:
Preference-based planners take these preferences into account when producing a plan for a given problem. Examples of preference-based planning software include
613:
291:
77:
559:
28:
193:
549:
618:
156:
This example metric function specifies that the calculated value of the plan should be minimized (i.e., a plan with value
252:
Deterministic planning in the fifth international planning competition: PDDL3 and experimental evaluation of the planners
539:
455:
369:
268:
196:, flexible variants exist that deal with soft constraints in a similar way to preferences in preference-based planning.
275:
172:, the former plan is strictly preferred). The value of a plan is computed by the given function, which is expressed in
435:
412:
364:
392:
98:
should be clean at each state of the plan. In other words, the planner should not schedule an action that causes
48:
496:
526:
20:
224:
577:
534:
511:
491:
397:
102:
to become dirty. As this example shows, a preference is evaluated with regard to all states of a plan (if
460:
341:
326:
316:
114:
592:
572:
402:
311:
374:
295:
481:
587:
544:
516:
427:
407:
306:
384:
331:
321:
61:
567:
336:
236:
173:
260:
68:
to school. This increases the length of the plan but the user's preference is satisfied.
607:
351:
251:
31:
which focuses on producing plans that additionally satisfy as many user-specified
417:
32:
103:
212:
151:(:metric minimize (+ (* 5 (is-violated pref1)) (* 7 (is-violated pref2))))
129:(the preference holds during at most one sequence of states in the plan).
501:
445:
80:, as of version 3.0, supports the specification of preferences through
476:
506:
486:
450:
359:
440:
264:
180:, has been given a greater penalty than the first preference,
176:. In this case, violation of the second preference,
558:
525:
469:
426:
383:
350:
237:Planning with Preferences using Logic Programming
276:
125:(to be planned after a particular state) and
8:
283:
269:
261:
16:Form of automated planning and scheduling
247:
245:
205:
84:statements. For example, the statement
60:These hard constraints are part of the
94:indicates that the user prefers that
7:
89:(preference (always (clean room1)))
78:Planning Domain Definition Language
72:Planning Domain Definition Language
14:
614:Automated planning and scheduling
121:(at least once during the plan),
29:automated planning and scheduling
194:constraint satisfaction problems
188:Constraints satisfaction problem
1:
225:HTN Planning with Preferences
540:Constraint logic programming
456:Knowledge Interchange Format
413:Procedural reasoning systems
370:Expert systems for mortgages
365:Connectionist expert systems
117:are also supported, such as
113:, other constructs based on
436:Attempto Controlled English
635:
583:Preference-based planning
302:
49:hierarchical task network
25:preference-based planning
292:Knowledge representation
140:is-violated <name>
527:Constraint satisfaction
21:artificial intelligence
578:Partial-order planning
535:Constraint programming
160:and a plan with value
461:Web Ontology Language
403:Deductive classifiers
342:Knowledge engineering
327:Model-based reasoning
317:Commonsense reasoning
115:linear temporal logic
619:Strategic management
593:State space planning
573:Multi-agent planning
375:Legal expert systems
312:Case-based reasoning
560:Automated planning
428:Ontology languages
398:Constraint solvers
239:, Son and Pontelli
47:(preference-based
601:
600:
588:Reactive planning
545:Local consistency
385:Reasoning systems
332:Inference engines
307:Backward chaining
254:, Gerevini et al.
215:, Bienvenu et al.
51:(HTN) planning).
626:
337:Proof assistants
322:Forward chaining
285:
278:
271:
262:
255:
249:
240:
234:
228:
227:, Sohrabi et al.
222:
216:
210:
183:
179:
152:
145:
141:
128:
124:
120:
112:
101:
97:
90:
83:
62:domain knowledge
634:
633:
629:
628:
627:
625:
624:
623:
604:
603:
602:
597:
568:Motion planning
554:
521:
470:Theorem provers
465:
422:
393:Theorem provers
379:
346:
298:
289:
259:
258:
250:
243:
235:
231:
223:
219:
211:
207:
202:
192:In the area of
190:
181:
177:
174:Polish notation
150:
143:
139:
135:
126:
122:
118:
110:
109:In addition to
99:
95:
88:
81:
74:
57:
17:
12:
11:
5:
632:
630:
622:
621:
616:
606:
605:
599:
598:
596:
595:
590:
585:
580:
575:
570:
564:
562:
556:
555:
553:
552:
547:
542:
537:
531:
529:
523:
522:
520:
519:
514:
509:
504:
499:
494:
489:
484:
479:
473:
471:
467:
466:
464:
463:
458:
453:
448:
443:
438:
432:
430:
424:
423:
421:
420:
415:
410:
408:Logic programs
405:
400:
395:
389:
387:
381:
380:
378:
377:
372:
367:
362:
356:
354:
352:Expert systems
348:
347:
345:
344:
339:
334:
329:
324:
319:
314:
309:
303:
300:
299:
290:
288:
287:
280:
273:
265:
257:
256:
241:
229:
217:
204:
203:
201:
198:
189:
186:
154:
153:
134:
131:
123:sometime-after
92:
91:
73:
70:
56:
53:
15:
13:
10:
9:
6:
4:
3:
2:
631:
620:
617:
615:
612:
611:
609:
594:
591:
589:
586:
584:
581:
579:
576:
574:
571:
569:
566:
565:
563:
561:
557:
551:
548:
546:
543:
541:
538:
536:
533:
532:
530:
528:
524:
518:
515:
513:
510:
508:
505:
503:
500:
498:
495:
493:
490:
488:
485:
483:
480:
478:
475:
474:
472:
468:
462:
459:
457:
454:
452:
449:
447:
444:
442:
439:
437:
434:
433:
431:
429:
425:
419:
416:
414:
411:
409:
406:
404:
401:
399:
396:
394:
391:
390:
388:
386:
382:
376:
373:
371:
368:
366:
363:
361:
358:
357:
355:
353:
349:
343:
340:
338:
335:
333:
330:
328:
325:
323:
320:
318:
315:
313:
310:
308:
305:
304:
301:
297:
293:
286:
281:
279:
274:
272:
267:
266:
263:
253:
248:
246:
242:
238:
233:
230:
226:
221:
218:
214:
209:
206:
199:
197:
195:
187:
185:
175:
171:
167:
163:
159:
149:
148:
147:
132:
130:
116:
107:
105:
87:
86:
85:
79:
71:
69:
65:
63:
54:
52:
50:
46:
42:
37:
34:
30:
27:is a form of
26:
22:
582:
418:Rule engines
232:
220:
208:
191:
169:
165:
161:
157:
155:
136:
133:Plan quality
127:at-most-once
108:
104:semantically
93:
75:
66:
58:
44:
40:
38:
24:
18:
550:SMT solvers
106:required).
33:preferences
608:Categories
200:References
164:such that
82:preference
296:reasoning
45:HTNPlan-P
119:sometime
55:Overview
502:Prover9
497:Paradox
446:F-logic
144::metric
477:CARINE
111:always
507:SPASS
492:Otter
487:Nqthm
451:FO(.)
360:CLIPS
213:PPLAN
182:pref1
178:pref2
168:<
100:room1
96:room1
41:PPLAN
441:CycL
294:and
76:The
43:and
512:TPS
19:In
610::
517:Z3
244:^
184:.
170:v2
166:v1
162:v2
158:v1
146::
23:,
482:E
284:e
277:t
270:v
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.