最长上升子序列

寻找数组中的最长递增子序列

两个思路,

第一个,所求子序列,必然以某一个元素结尾,求出所有元素结尾的最长子序列,挑一个最大的,就可以了。辅助数组存的是对应下标的元素的最长值。

第二个,辅助数组存的是对应长度的最小值,可以反证法证明,长度越大的最小元素值,大于长度比他小的最小元素值。因此,该辅助数组是单调递增的,可以二分,提高效率。


public class LIS {

	private int[] a;
	private int[] b;
	
	public LIS(int[] a){
		assert(a != null);
		assert(a.length>0);
		this.a = a;
		b = new int[a.length];
	}
	
	/**
	 * The max subsequence must be ended with one(or several) of the array,  
	 * To find the max subsequence, is to find all the max subsequences ending with each array member, and choose
	 * the max value in them. That is the result.
	 * @return
	 */
	public int find(){
		//the max counts of subsequence equals the b.length.
		int max = 1;
		/*for the a[i], to generate the max sub ending with a[i], which item appeared before 
		should the a[i] follow? 
		so we need an array to record all the happened counts.
		b[n] stands the min ending value in all the possible max subsequences 
		happened before whose length are all (n-1). 
		why is ONLY the min ending value saved? because for the next following item, the ending value
		 smaller means the new item will easier to follow to generate a longer subsequence. 
		 b has 2 characteristic.
		1) the index is the max length.
		2) b is sorted well naturally.
		*/
		b[0] = a[0];
		//find the max sub each one by one, ending with a[i]
		for(int i=1; i<a.length-1; i++){
			//find the right place in array b, make the subsequence ending with a[i] has the max length.
			int position = search(a[i], b, 0, max-1);
			//if exceed the max, add it directly.
			if(position == (max-1)){
				b[max] = a[i];
				max++;
			}
			if(a[i] < b[position+1]){
				b[position+1] = a[i];
			}
		}
//		print(b, 0, max-1);
		return max;
	}
	
	public int find2(){
		for(int i=0; i<a.length; i++){
			int max = 1;
			for(int j=0; j<i; j++){
				if(a[j] < a[i]){
					int temp = b[j] + 1;
					if(temp > max){
						max = temp;
					}
				}
			}
			b[i] = max;
		}
		int result = 1;
		for(int k=0; k<b.length; k++){
			if(b[k] > result){
				result = b[k];
			}
		}
		return result;
	}
	/**
	 * find the right place in array b, make the subsequence ending with a[i] has the max length.
	 * @param value
	 * @param b
	 * @return
	 */
	public int search(int value, int[] b, int left, int right){
		assert(b != null);
		if(value <= b[left]){
			return -1;
		}
		if(value > b[right]){
			return right;
		}
		return binarySearch(value, b, left, right);
	}
	/**
	 * Find the closest item in array b which is less than the value.  and return its index.
	 * @param value
	 * @return 
	 */
	private int binarySearch(int value, int[] b, int left, int right){
		//why always return "right"?
		if(left > right){
			return right;
		}
		int index = (left+right)/2;
		int mid = b[index];
		if(mid == value){
			return index-1;
		}else if(value > mid){
			return binarySearch(value, b, index+1, right);
		}else{
			return binarySearch(value, b, left, index-1);
		}
	}
	
	public void print(int[] b, int left, int right){
		for(int i=left; i<=right; i++){
			System.out.print(b[i] + " ");
		}
		System.out.print("\r");
	}
	/**
	 * @param args
	 */
	public static void main(String[] args) {
		int[] a = {7,2,4,8,3,5,1,9,6};
		LIS seq = new LIS(a);
		int max = seq.find();
//		int max = seq.find2();
		System.out.println(max);
		
		
	}

}


内容概要:本文介绍了一个基于冠豪猪优化算法(CPO)的无人机三维路径规划项目,利用Python实现了在复杂三维环境中为无人机规划安全、高效、低能耗飞行路径的完整解决方案。项目涵盖空间环境建模、无人机动力学约束、路径编码、多目标代价函数设计以及CPO算法的核心实现。通过体素网格建模、动态障碍物处理、路径平滑技术和多约束融合机制,系统能够在高维、密集障碍环境下快速搜索出满足飞行可行性、安全性与能效最优的路径,并支持在线重规划以适应动态环境变化。文中还提供了关键模块的代码示例,包括环境建模、路径评估和CPO优化流程。; 适合人群:具备一定Python编程基础和优化算法基础知识,从事无人机、智能机器人、路径规划或智能优化算法研究的相关科研人员与工程技术人员,尤其适合研究生及有一定工作经验的研发工程师。; 使用场景及目标:①应用于复杂三维环境下的无人机自主导航与避障;②研究智能优化算法(如CPO)在路径规划中的实际部署与性能优化;③实现多目标(路径最短、能耗最低、安全性最高)耦合条件下的工程化路径求解;④构建可扩展的智能无人系统决策框架。; 阅读建议:建议结合文中模型架构与代码示例进行实践运行,重点关注目标函数设计、CPO算法改进策略与约束处理机制,宜在仿真环境中测试不同场景以深入理解算法行为与系统鲁棒性。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值