string temp = @"比如说我有个网页内容如下
<link type='text/css' rel='stylesheet' href='a.css' /><link type='text/css' rel='stylesheet' href='b.css' /><div><input/></div>
 
";
                string[] arr= Regex.Matches(temp, @"(?i)<link[^>]*?href=(['""])([^'""]*?)\1[^>]*?>").Cast<Match>().Select(a => a.Groups[2].Value).ToArray();
               /*
                * a.css
                * b.css
                */

------解决方案--------------------
 string str = "<link type='text/css' rel='stylesheet' href='a.css' /><link type='text/css' rel='stylesheet' href='b.css' /><div><input/></div>";
           Regex.Matches(str, @"<link[\s\S]*?href=['""]([^'""]+)['""][^/]+/>").Cast<Match>().ToList().ForEach(t => Console.WriteLine(t.Groups[1].Value));


------解决方案--------------------
(?i)<link\b[^>]*?href=(['"]?)([^"']*?)\1[^>]*?/>

取Groups[2].Value