最快的技术来存储在android文件中存储double [] []。将double [] []连接到字符串

Question

我的代码合并了两个android传感器数据double [] []数组，然后连接成一个文件存储字符串。合并代码对齐时间戳并且工作正常，但字符串连接对于大型数据阵列（例如30,000条记录）来说非常慢。有没有更好的方法来将double [] []数组存储为文件中的字符串数据？在性能方面，是否有更好的创建字符串对象的技术？性能问题出现在下面的for循环的最后一组中。

package com.xxxx.android.sara.dva;

import java.lang.reflect.Array;

public final class CombineDblArr {
public double[][] rtrnArr;
public String bigFileStrng;

CombineDblArr(double[][] lrgArr, int lrgTSDim, double[][] smlArr,
        int smlTSDim, double startTS) {
    // find the closest TSes for each array
    // pull the data from the arrays into a file loading string
    // search the arrays for the first timestamp that is nearest the
    // start TS
    double d = -1;
    double bestDistanceNow = Double.MAX_VALUE;
    int locLrgDex = 0;
    int locSmlDex = 0;
    int lrgSize = lrgArr.length;
    for (int i = 0; i < lrgSize; i++) {
        if (lrgArr[i][lrgTSDim] == startTS) {
            locLrgDex = i;
            break;
        } else {
            d = Math.abs(lrgArr[i][lrgTSDim] - startTS);
            // cannot optimize the search due to cyclical TS design of
            // arrays
            // the max TS delta could be in consecutive rows
            if (d < bestDistanceNow) {
                bestDistanceNow = d;
                locLrgDex = i;
            }
        }
    }
    d = -1;
    bestDistanceNow = Double.MAX_VALUE;
    int smlSize = smlArr.length;
    for (int i = 0; i < smlSize; i++) {
        if (smlArr[i][smlTSDim] == startTS) {
            locSmlDex = i;
            break;
        } else {
            d = Math.abs(smlArr[i][smlTSDim] - startTS);
            // cannot optimize the search due to cyclical TS design of
            // arrays
            // the max TS delta could be in consecutive rows
            if (d < bestDistanceNow) {
                bestDistanceNow = d;
                locSmlDex = i;
            }
        }
    }
    // get the dims of the 2D arrays
    int lrgDmsn = 0;
    int smlDmsn = 0;
    try {
        lrgDmsn = lrgArr[0].length;
        smlDmsn = smlArr[0].length;
    } catch (ArrayIndexOutOfBoundsException e) {
        // TODO Auto-generated catch block
        e.printStackTrace();
    } catch (IllegalArgumentException e) {
        // TODO Auto-generated catch block
        e.printStackTrace();
    }
    rtrnArr = new double[lrgSize][lrgDmsn + smlDmsn];
    for (int j = 0; j < lrgSize; j++) {

        // align the TS when combining rows. both arrays are the same size,
        // but diff TS values
        for (int k = 0; k < lrgDmsn; k++) {
            rtrnArr[j][k] = lrgArr[locLrgDex][k];
        }
        for (int k = 0; k < smlDmsn; k++) {
            rtrnArr[j][k + lrgDmsn - 1] = smlArr[locSmlDex][k];
        }
        // see if the next TS in smlArr is within range of the next TS in
        // lrgArr
        locLrgDex++;
        // check for rollover condition
        if ((locLrgDex + 1) >= lrgSize)
            locLrgDex = 0;
        if (smlArr[locSmlDex + 1][smlTSDim] < lrgArr[locLrgDex + 1][lrgTSDim])
            locSmlDex++;
        if (locSmlDex >= lrgSize)
            locSmlDex = 0;
    }
    // create the writeable string
    // this code is incredibly slow
    int bigFileDim = lrgDmsn + smlDmsn;
    bigFileStrng = "";
    for(int l=0; l<lrgSize;l++){
        // break if a TS value is 0
        if(rtrnArr[l][lrgTSDim] != 0.0){
            // build the string
            for (int m = 0; m < bigFileDim; m++)
                bigFileStrng = bigFileStrng + (Double.toString(rtrnArr[l][m])) + ",";
            bigFileStrng = bigFileStrng + "\n";
        }else
            break;
    }
    return;
}

}