{"id":235,"date":"2018-07-12T09:57:07","date_gmt":"2018-07-12T12:57:07","guid":{"rendered":"http:\/\/www.professores.uff.br\/lbertini\/?page_id=235"},"modified":"2018-07-12T09:57:07","modified_gmt":"2018-07-12T12:57:07","slug":"controlador-fuzzy-em-c","status":"publish","type":"page","link":"https:\/\/www.professores.uff.br\/lbertini\/controlador-fuzzy-em-c\/","title":{"rendered":"Controlador Fuzzy em C"},"content":{"rendered":"

O controlador apresentado abaixo foi testado no controle de velocidade de um motor BLDC de CDROM, no controle de velocidade. Foi comparado com um controlador PID implementado utilizando transformada Z e testado com diferentes par\u00e2metros de integra\u00e7\u00e3o. O resultado do controle e um caso de ativa\u00e7\u00e3o de regras podem ser vistos nas figuras anbaixo:<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

 <\/p>\n

C\u00f3digo:<\/strong><\/p>\n

<\/p>\n

\n
\/*<\/span>\r\nCodigo de um controlador Fuzzy com dois precedentes conectados pelo operador AND.<\/span>\r\nSISTEMAS EMBARCADOS E COMPUTA\u00c7\u00c3O UB\u00cdQUA<\/span>\r\nMESTRADO EM ENGENHARIA DE PRODU\u00c7\u00c3O E SISTEMAS COMPUTACIONAIS \u2013 MESC<\/span>\r\nLuciano Bertini *\/<\/span>\r\n\r\n\r\n#include <stdio.h><\/span>\r\n\r\n\/\/ Define a estrutura para um conjunto fuzzy triangular, sendo que a ou c podem ser +infinito ou -infinito (INF)<\/span>\r\nstruct<\/span> membership{\r\n    float<\/span> a;\r\n    float<\/span> b;\r\n    float<\/span> c;\r\n};\r\n\r\n\/\/Defini\u00e7\u00e3o de infinito<\/span>\r\n#define INF 1e+6<\/span>\r\n\r\n#define NUM1 5  <\/span>\/\/Numero de conjuntos na variavel de entrada 1<\/span>\r\n#define NUM2 5  <\/span>\/\/Numero de conjuntos na variavel de entrada 2<\/span>\r\n#define NUM3 5  <\/span>\/\/Numero de conjuntos na variavel de saida<\/span>\r\n\r\n#define MIN(a,b) (((a)<(b))?(a):(b))<\/span>\r\n#define MAX(a,b) (((a)>(b))?(a):(b))<\/span>\r\n\r\n\/\/ Fun\u00e7\u00e3o membership de entrada. r[] \u00e9 o vetor de pertinencias que ser\u00e1 preenchido.<\/span>\r\n\/\/ x \u00e9 o valor sendo fuzificado. z[] \u00e9 o vetor de conjuntos fuzzy em todo universo de discurso X<\/span>\r\n\/\/ n \u00e9 o tamanho do vetor<\/span>\r\nvoid<\/span> test_membership_in<\/span>(float<\/span> r[], float<\/span> x, struct<\/span> membership z[], int<\/span> n)\r\n{\r\n    int<\/span> i;\r\n    for<\/span>(i=<\/span>0<\/span>;i<<\/span>n;i++<\/span>)\r\n    {\r\n         if<\/span>(x<=<\/span>z[i].a||<\/span>x>=<\/span>z[i].c)\r\n            r[i] =<\/span> 0.0<\/span>;\r\n         else<\/span> if<\/span>(x><\/span>z[i].a&&<\/span>x<=<\/span>z[i].b)\r\n            r[i] =<\/span> (x-<\/span>z[i].a)\/<\/span>(z[i].b-<\/span>z[i].a);\r\n         else<\/span> if<\/span>(x><\/span>z[i].b&&<\/span>x<<\/span>z[i].c)\r\n            r[i] =<\/span> (z[i].c-<\/span>x)\/<\/span>(z[i].c-<\/span>z[i].b);\r\n    }\r\n}\r\n\r\n\r\n\/\/ Obtem o valor da membership de saida ap\u00f3s a infer\u00eancia fuzzy. O vetor r[] \u00e9 entrada e cada valor \u00e9 <\/span>\r\n\/\/ o n\u00edvel de ativa\u00e7\u00e3o de cada conjunto fuzzy, obtido na inferencia. Variando-se x em todo universo de discurso, <\/span>\r\n\/\/ obtem-se a membership de saida<\/span>\r\nfloat<\/span> test_membership_out<\/span>(float<\/span> r[],float<\/span> x, struct<\/span> membership z[], int<\/span> n)\r\n{\r\n    float<\/span> v, ret=<\/span>0.0<\/span>;\r\n    int<\/span> i;\r\n    for<\/span>(i=<\/span>0<\/span>;i<<\/span>n;i++<\/span>)\r\n    {\r\n         if<\/span>(x<=<\/span>z[i].a||<\/span>x>=<\/span>z[i].c)\r\n            v =<\/span> 0.0<\/span>;\r\n         else<\/span> if<\/span>(x><\/span>z[i].a&&<\/span>x<=<\/span>z[i].b)\r\n            v =<\/span> MIN((x-<\/span>z[i].a)\/<\/span>(z[i].b-<\/span>z[i].a),r[i]); \/\/ seleciona o minimo entre o nivel de ativa\u00e7\u00e3o e o triangulo do conjunto fuzzy<\/span>\r\n         else<\/span> if<\/span>(x><\/span>z[i].b&&<\/span>x<<\/span>z[i].c)\r\n            v =<\/span> MIN((z[i].c-<\/span>x)\/<\/span>(z[i].c-<\/span>z[i].b),r[i]); \/\/ seleciona o minimo entre o nivel de ativa\u00e7\u00e3o e o triangulo do conjunto fuzzy<\/span>\r\n         ret=<\/span>MAX(ret,v); \/\/ O maximo aqui faz o OU entre dois conjuntos fuzzy i-1 e i<\/span>\r\n    }\r\n    return<\/span> ret;\r\n}\r\nmain()\r\n{\r\n    int<\/span> i,j;\r\n    float<\/span> de=<\/span>1.7<\/span>, e=-<\/span>1.4<\/span>;\r\n    float<\/span> x,sum1=<\/span>0.0<\/span>,sum2=<\/span>0.0<\/span>, val, v;\r\n    float<\/span> mi1[NUM1],mi2[NUM2],out[NUM3]=<\/span>{0<\/span>,0<\/span>,0<\/span>,0<\/span>,0<\/span>};\r\n\r\n   \/\/ Defini\u00e7\u00e3o dos conjuntos fuzzy<\/span>\r\n   struct<\/span> membership e1[NUM1]=<\/span>{\r\n        { -<\/span>INF, -<\/span>8.0<\/span>, -<\/span>4.0<\/span>},\r\n        { -<\/span>5.0<\/span>, -<\/span>3.0<\/span>, -<\/span>1.0<\/span>},\r\n        { -<\/span>2.0<\/span>,  0.0<\/span>,  2.0<\/span>},\r\n        {  1.0<\/span>,  3.0<\/span>,  5.0<\/span>},\r\n        {  4.0<\/span>,  8.0<\/span>,  INF}};\r\n\r\n    struct<\/span> membership e2[NUM2]=<\/span>{\r\n        { -<\/span>INF, -<\/span>8.0<\/span>, -<\/span>4.0<\/span>},\r\n        { -<\/span>5.0<\/span>, -<\/span>3.0<\/span>, -<\/span>1.0<\/span>},\r\n        { -<\/span>2.0<\/span>,  0.0<\/span>,  2.0<\/span>},\r\n        {  1.0<\/span>,  3.0<\/span>,  5.0<\/span>},\r\n        {  4.0<\/span>,  8.0<\/span>,  INF}};\r\n\r\n    struct<\/span> membership s[NUM3]=<\/span>{\r\n        {-<\/span>20.0<\/span>, -<\/span>10.0<\/span>, -<\/span>6.0<\/span>},\r\n        { -<\/span>7.0<\/span>,  -<\/span>4.0<\/span>, -<\/span>1.0<\/span>},\r\n        { -<\/span>2.0<\/span>,   0.0<\/span>,  2.0<\/span>},\r\n        {  1.0<\/span>,   4.0<\/span>,  7.0<\/span>},\r\n        {  6.0<\/span>,  10.0<\/span>, 20.0<\/span>}};\r\n\r\n\r\n    \/\/ Mapa de regras<\/span>\r\n    int<\/span> regras[NUM1][NUM2]=<\/span>{ \/\/horizontal Delta do erro<\/span>\r\n        {4<\/span>,4<\/span>,3<\/span>,3<\/span>,2<\/span>},         \/\/vertical erro<\/span>\r\n        {4<\/span>,3<\/span>,3<\/span>,2<\/span>,1<\/span>},\r\n        {3<\/span>,3<\/span>,2<\/span>,1<\/span>,1<\/span>},\r\n        {3<\/span>,2<\/span>,1<\/span>,1<\/span>,0<\/span>},\r\n        {2<\/span>,1<\/span>,1<\/span>,0<\/span>,0<\/span>}};\r\n\r\n    \/\/ Inicializa\u00e7\u00e3o<\/span>\r\n    sum1=<\/span>0<\/span>;\r\n    sum2=<\/span>0<\/span>;\r\n    for<\/span>(i=<\/span>0<\/span>;i<<\/span>NUM3;i++<\/span>)out[i]=<\/span>0<\/span>;\r\n\r\n    \/\/ Fuzifica\u00e7\u00e3o<\/span>\r\n    test_membership_in(mi1,e,e1,NUM1);\r\n    test_membership_in(mi2,de,e2,NUM2);\r\n\r\n    \/\/ Inferencia Fuzzy<\/span>\r\n    \/\/ O minimo aqui pega o menor dos graus de pertinencias, supondo regras com dois precedentes conectados pelo operador AND<\/span>\r\n    \/\/ O m\u00e1ximo aqui \u00e9 util quando um certo conjunto fuzzy de sa\u00edda j\u00e1 foi ativado antes por outra regra e neste caso deve-se selecionar o maior (OR).<\/span>\r\n    for<\/span>(i=<\/span>0<\/span>;i<<\/span>NUM1;i++<\/span>)\r\n        for<\/span>(j=<\/span>0<\/span>;j<<\/span>NUM2;j++<\/span>)\r\n            if<\/span>(mi1[i]><\/span>0<\/span>&&<\/span>mi2[j]><\/span>0<\/span>)\r\n                out[regras[i][j]] =<\/span> MAX(MIN(mi1[i],mi2[j]), out[regras[i][j]]); \r\n\r\n    \/\/ Defuzifica\u00e7\u00e3o<\/span>\r\n    \/\/ Calculo do centroide<\/span>\r\n    for<\/span>(x=-<\/span>10.0<\/span>;x<=<\/span>10.01<\/span>;x+=<\/span>0.05<\/span>)\r\n    {\r\n        val=<\/span>test_membership_out(out, x, s, NUM3);\r\n        sum1+=<\/span>x*<\/span>val;\r\n        sum2+=<\/span>val;\r\n    }\r\n    x=<\/span>sum1\/<\/span>sum2;\r\n    printf(\"saida=%f<\/span>\\n<\/span>\"<\/span>,x);\r\n}\r\n<\/pre>\n<\/div>\n
 <\/p>\n
 <\/p>\n","protected":false},"excerpt":{"rendered":"
O controlador apresentado abaixo foi testado no controle de velocidade de um motor BLDC de CDROM, no controle de velocidade. Foi comparado com um controlador PID implementado utilizando transformada Z e testado com diferentes par\u00e2metros de integra\u00e7\u00e3o. O resultado do controle e um caso de ativa\u00e7\u00e3o de regras podem ser vistos nas figuras anbaixo:   […]<\/p>\n","protected":false},"author":168,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_exactmetrics_skip_tracking":false,"_exactmetrics_sitenote_active":false,"_exactmetrics_sitenote_note":"","_exactmetrics_sitenote_category":0,"footnotes":""},"categories":[],"tags":[],"class_list":["post-235","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/pages\/235","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/users\/168"}],"replies":[{"embeddable":true,"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/comments?post=235"}],"version-history":[{"count":1,"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/pages\/235\/revisions"}],"predecessor-version":[{"id":238,"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/pages\/235\/revisions\/238"}],"wp:attachment":[{"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/media?parent=235"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/categories?post=235"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.professores.uff.br\/lbertini\/wp-json\/wp\/v2\/tags?post=235"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}