文档介绍：Sｈiｐ Desｉgn Opｔimizatｉon
Ｔhis ｃoｎtribｕtiｏn ｉｓ devｏｔed to explｏiting tｈe analoｇｙ bｅｔwｅen a ｍｏdｅrｎｍａnufａｃtuｒｉｎgpｌant ａnd a hｅtｅｒｏgeneouｓ pａralｌｅｌ computeｒ ｔo ｃoｎstruct ａHPCN ｄｅｃisioｎ sｕｐport ｔooｌ fｏｒ shｉｐ dｅｓiｇnｅrs. Thｅ aｐpｌicatｉｏn ｉs a HPＣN ｏnebecａusｅ of the sｃale oｆ shｉｐｂuiｌｄinｇ － a ｌａrgｅ contaｉner vesseｌ is consｔrucｔed ｂyassｅmbｌiｎg aｂouｔ 1。５ ｍilｌion atomｉc coｍpｏnｅnts iｎ a pｒoｄuｃtｉｏn hｉｅrａrcｈｙ． Thｅｒolｅ of ｔｈe decision sｕpport tｏol iｓ tｏ rａpｉdｌｙ eｖalｕatｅ the manufａctｕringｃｏnｓeｑｕencｅｓ ｏf deｓign changｅs。 Ｔhe implｅｍentation aｓ a diｓtributed muｌｔi-agｅntaｐpliｃaｔiｏn ｒｕnning on ｔｏp ｏf PVM is dｅscribed
1 Anaｌogｉeｓ betweeｎ Ｍａnｕｆａctｕriｎｇ and HＰCN
Tｈere ａre a number oｆ analｏgies beｔwｅeｎ ｔhe ｍanuｆaｃｔｕrｅ ｏf ｃｏｍpleｘ prodｕcｔsｓuch as ｓhiｐs， ａircrａｆt ａnd cａrs anｄ tｈｅ executｉon ｏf a pａｒａllｅl program。 Tｈｅｍanｕfacｔｕre of a shiｐ ｉｓ carｒied out ａｃcordｉng ｔo a produｃtｉon pｌan wｈich enｓｕｒes tｈatａll the cｏmpoｎeｎtｓ cｏme ｔoｇethｅr at the rigｈt timｅ aｔ thｅ ｒight ｐlａｃe. A ｐarａllelcoｍpｕter apｐliｃaｔioｎ ｓhoｕｌd ensure tｈａt tｈｅ ａpprｏｐｒiａtｅ ｄａｔａ iｓ aｖａilａblｅ ｏn ｔｈeａpproprｉａtｅ pｒｏcｅssor in a tｉｍelｙ ｆashｉon．
Iｔ iｓ not surpｒｉsiｎg， theｒeｆoｒe, that mａnｕｆaｃtuｒiｎg iｓ ｐｌaｇued by ｉｎdeteｒｍinacyexacｔｌy ａｓ are paraｌｌeｌ programs exｅｃｕting on multｉ-proｃｅssor hardｗaｒe. Thｉs hａｓcａuｓed a numｂｅｒ of rｅsearｃｈｅrs in ｐroｄuctiｏn enｇｉneｅｒiｎg ｔo seek ｉｎspiraｔｉon ｉｎ otheｒａｒeａｓ ｗhｅｒe ｍanaging coｍｐlexitｙ ａnd unｐrｅｄicｔabiｌity is importaｎt。 A nｕmber oｆ ｎｅｗpａｒadｉｇmｓ, ｓucｈ aｓ Holoｎiｃ Mａnufａctuｒｉng ａnd Fractal Ｆａcｔoｒｉes havｅ ｅmｅrｇed ［1，2]ｗhicｈ ｃontain iｄｅas rather rｅｍinｉsceｎｔ of tｈｏｓe to ｂｅ fｏund in tｈe ｆiｅlｄ ｏf Mｕｌｔi-Ageｎt Systｅｍｓ ［３， 4］.
Mａｎｕfａctｕring taｓkｓ arｅ analogous to operatiｏns carｒieｄ ouｔ on daｔａ, wｉｔhin ｔｈeｃｏｎｔext ｏf plaｎnｉnｇ， schedulｉnｇ and control． Also， compleｘ ｐroｄｕcｔs aｒe assｅｍｂlｅｄ atphyｓiｃalｌｙ ｄｉstributｅd workshops or pｒoduｃtiｏn facｉlｉtiｅｓ, so thｅ ｃｏｍｐonｅｎts mｕｓｔ ｂｅtransporteｄ beｔween ｔhem。 Ｔｈis is analｏgous ｔｏ cｏｍｍｕｎicatioｎ ｏf dａｔa betｗeenpｒoｃessors ｉn a paｒａｌｌel compuｔer， ｗｈich tｈuｓ ａlsｏ mａkｅs ｃlear the aｎalｏgｙ