You can change this behaivor by attaching your self to  this event

UrlSegment.CreatingUrlSegment += new EventHandler<UrlSegmentEventArgs>(UrlSegment_CreatingUrlSegment);

    Then you can change the replace string, or as this example translate to english when characters in the url segment is only -

       static void UrlSegment_CreatingUrlSegment(object sender, UrlSegmentEventArgs e)
        {
            string uRLSegment = e.PageData.URLSegment;
            if (string.IsNullOrEmpty(uRLSegment))
            {
                uRLSegment = e.PageData.PageName;
                if (uRLSegment == null)
                {
                    uRLSegment = "";
                }
            }
            string urlFriendlySegment = ReplaceIllegalChars(uRLSegment);
            if (string.IsNullOrEmpty(urlFriendlySegment) || urlFriendlySegment.Replace("-","")=="")
            {
                urlFriendlySegment=TranslateUsingGoogle(e.PageData.PageName, e.PageData.LanguageID, "en");
            }
            e.PageData.URLSegment = urlFriendlySegment;
           
        }
        internal static string ReplaceIllegalChars(string inputString)
        {
            string InvalidSegmentNames = @"%|^COM[0-9]([/\.]|$)|^LPT[0-9]([/\.]|$)|^PRN([/\.]|$)|^CLOCK\$([/\.]|$)|^AUX([/\.]|$)|^NUL([/\.]|$)|^CON([/\.]|$)";
            Regex  regexValidUrlChars = new Regex(@"^[A-Za-z0-9\-_~]+$", RegexOptions.Compiled);
            Regex regexFindInvalidUrlChars = new Regex(@"[^A-Za-z0-9\-_~]{1}", RegexOptions.Compiled);
            Regex regexInvalidSegmentNames = new Regex(InvalidSegmentNames, RegexOptions.Compiled | RegexOptions.IgnoreCase);


            StringBuilder builder = new StringBuilder(inputString);
            MatchCollection matchs = regexFindInvalidUrlChars.Matches(inputString);
            for (int i = 0; i < matchs.Count; i++)
            {
                object obj2 = UrlSegment.GetURLCharacterMap()[builder[matchs[i].Index]];
                if (obj2 != null)
                {
                    builder[matchs[i].Index] = (char)obj2;
                }
                else
                {
                    builder[matchs[i].Index] = '?';
                }
            }
            builder.Replace("?", "");
            return builder.ToString();
        }

public static string TranslateUsingGoogle(string text, string fromLang, string toLang)
        {
            if (fromLang == null)
                fromLang = "auto";
            if (toLang == null)
                toLang = "en";

            string address = string.Format("http://www.google.com/translate_t?hl=en&ie=UTF8&text={0}&langpair={1}", text, fromLang + "|" + toLang);
            string str7 = new WebClient().DownloadString(address);
            str7 = str7.Substring(str7.IndexOf("id=result_box"), 500);
            str7 = str7.Substring(str7.IndexOf(">"));
            str7 = str7.Substring(0, str7.IndexOf("</div"));
            Regex removeSpan=new Regex("<[^>]*>");
            str7=removeSpan.Replace(str7,"");
           
            return str7;
        }

Thanks for the reply, but it doesn’t recognise the characters that are causing my issues, to replaces them with nothing, as in the build in functionality.

Is it possible to add items to the GetURLCharacterMap() collection?

you have to make your own CreatingUrlSegment, and for instance replace ě" and the "č to e and c before you call the base method