php utf-8不起作用

I'm trying, put in header("Content-Type: text/html; charset=utf-8"); but not work

<?PHP
  header("Content-Type: text/html; charset=utf-8");
  function randStr($rts=20){ 
  $act_chars = "ABCÇDEFGĞHIİJKLMNOÖPRSŞTUÜVYZ"; 
  $act_val = ""; 
  for($act=0; $act <$rts ; $act++) 
  { 
     mt_srand((double)microtime()*1000000); 
     $act_val .= $act_chars[mt_rand(0,strlen($act_chars)-1)]; 
  } 

  return $act_val; 
  } 
  $dene = randStr(16);
  print "$dene";
  ?>

output

K��A�CÞZU����EJ

There are several mistakes. Don't use [] or strlen() on a multibyte string. Use mb_substr() and mb_strlen().

Also set the internal encoding of the multibyte extension to UTF-8:

mb_internal_encoding("UTF-8");

You could also set the encoding on a per-function basis. See the specific function signatures for further details.

Here is an improved version of your code:

header('Content-Type: text/html; charset=utf-8');
mb_internal_encoding("UTF-8");

function randStr($rts = 20) { 
  $act_chars = 'ABCÇDEFGĞHIİJKLMNOÖPRSŞTUÜVYZ';
  $act_val = '';
  $act_chars_last = mb_strlen($act_chars);

  for($act = 0; $act < $rts; $act++) { 
    $act_val .= mb_substr($act_chars, mt_rand(0, $act_chars_last), 1);
  } 

  return $act_val; 
} 
$dene = randStr(16);
print $dene;

Replaced double quotes by single quotes (saves a tiny amount of time)
Removed mt_srand() calls because they are automatically done as of PHP 4.2.0 (thanks to feeela for mentioning that in a comment)
Saved the string length - 1 in a variable, so PHP doesn't need to recomputed it in every loop step.

If you want more insights on that topic, checkout the link from deceze in the answer below: What Every Programmer Absolutely, Positively Needs To Know About Encodings And Character Sets To Work With Text.

The problem is that you're assembling a random gobbledygook of bytes and are telling the browser to interpret it as UTF-8. The standard PHP str functions assume one byte = one character. By randomly picking a multi-byte string apart with those you're not going to get whole characters, only bytes.

If you need encoding aware functions operating on a character level rather than a byte level, use the mb_* functions.

And read What Every Programmer Absolutely, Positively Needs To Know About Encodings And Character Sets To Work With Text.

Instead of using the mb_* functions, you could alternatively set the following variables in php.ini if you have access to it:

# following mbstring-variables should be set via php.ini or vhost-configuration (httpd.conf);
# does not work per directory/via .htaccess
mbstring.language = Neutral
mbstring.internal_encoding = UTF-8
mbstring.func_overload = 7

That way you could use your current function as it is and use strlen or similar functions as if they were a mb_strlen function.

See php.net – Function Overloading Feature for more details.