中文 ZH
在 GitHub 上查看

Kalman Filter Trend Strategy

此趋势跟随方法使用卡尔曼滤波器平滑价格波动，并估计潜在方向。滤波器可动态适应市场噪声，比常规均线更能反映趋势力度。

测试表明年均收益约为 112%，该策略在外汇市场表现最佳。

当收盘价上穿卡尔曼估算值时做多；收盘价下穿时做空。由于滤波器每根K线更新，价格穿越即转仓，可持续参与趋势行情。ATR止损防止趋势突然反转带来的损失。

细节

  • 入场条件:
    • 多头: Close > Kalman Filter
    • 空头: Close < Kalman Filter
  • 多/空: 双向
  • 离场条件:
    • 多头: 收盘价跌破卡尔曼滤波
    • 空头: 收盘价升破卡尔曼滤波
  • 止损: 基于ATR
  • 默认值:
    • ProcessNoise = 0.01m
    • MeasurementNoise = 0.1m
    • CandleType = TimeSpan.FromMinutes(5)
  • 过滤器:
    • 类别: Trend
    • 方向: 双向
    • 指标: Kalman Filter
    • 止损: 是
    • 复杂度: 中等
    • 时间框架: 日内
    • 季节性: 否
    • 神经网络: 否
    • 背离: 否
    • 风险等级: 中等
using System;
using System.Linq;
using System.Collections.Generic;

using Ecng.Common;
using Ecng.Collections;
using Ecng.Serialization;

using StockSharp.Algo.Indicators;
using StockSharp.Algo.Strategies;
using StockSharp.BusinessEntities;
using StockSharp.Messages;

namespace StockSharp.Samples.Strategies;

/// <summary>
/// Kalman Filter Trend strategy.
/// Uses a custom Kalman Filter indicator to track price trend.
/// </summary>
public class KalmanFilterTrendStrategy : Strategy
{
	private readonly StrategyParam<decimal> _processNoiseParam;
	private readonly StrategyParam<decimal> _measurementNoiseParam;
	private readonly StrategyParam<DataType> _candleTypeParam;

	private KalmanFilter _kalmanFilter;
	private AverageTrueRange _atr;

	/// <summary>
	/// Process noise coefficient for Kalman filter.
	/// </summary>
	public decimal ProcessNoise
	{
		get => _processNoiseParam.Value;
		set => _processNoiseParam.Value = value;
	}

	/// <summary>
	/// Measurement noise coefficient for Kalman filter.
	/// </summary>
	public decimal MeasurementNoise
	{
		get => _measurementNoiseParam.Value;
		set => _measurementNoiseParam.Value = value;
	}

	/// <summary>
	/// Candle type for strategy.
	/// </summary>
	public DataType CandleType
	{
		get => _candleTypeParam.Value;
		set => _candleTypeParam.Value = value;
	}

	/// <summary>
	/// Constructor.
	/// </summary>
	public KalmanFilterTrendStrategy()
	{
		_processNoiseParam = Param(nameof(ProcessNoise), 0.01m)
			.SetRange(0.0001m, 1)
			.SetDisplay("Process Noise", "Process noise coefficient for Kalman filter", "Parameters")
			
			.SetOptimize(0.001m, 0.1m, 0.005m);

		_measurementNoiseParam = Param(nameof(MeasurementNoise), 0.1m)
			.SetRange(0.0001m, 1)
			.SetDisplay("Measurement Noise", "Measurement noise coefficient for Kalman filter", "Parameters")
			
			.SetOptimize(0.01m, 1.0m, 0.1m);

		_candleTypeParam = Param(nameof(CandleType), TimeSpan.FromMinutes(5).TimeFrame())
			.SetDisplay("Candle Type", "Candle type for strategy", "Common");
	}

	/// <inheritdoc />
	public override IEnumerable<(Security sec, DataType dt)> GetWorkingSecurities()
	{
		return [(Security, CandleType)];
	}

	/// <inheritdoc />
	protected override void OnReseted()
	{
		base.OnReseted();

		_kalmanFilter = null;
		_atr = null;
	}

	/// <inheritdoc />
	protected override void OnStarted2(DateTime time)
	{
		base.OnStarted2(time);

		// Create indicators
		_kalmanFilter = new KalmanFilter 
		{ 
			ProcessNoise = ProcessNoise,
			MeasurementNoise = MeasurementNoise 
		};
		
		_atr = new AverageTrueRange { Length = 14 };

		// Create subscription and bind indicators
		var subscription = SubscribeCandles(CandleType);
		subscription
			.Bind(_kalmanFilter, _atr, ProcessCandle)
			.Start();

		// Setup chart visualization if available
		var area = CreateChartArea();
		if (area != null)
		{
			DrawCandles(area, subscription);
			DrawIndicator(area, _kalmanFilter);
			DrawOwnTrades(area);
		}
		
		// Enable position protection
		StartProtection(
			takeProfit: new Unit(0, UnitTypes.Absolute), // No take profit
			stopLoss: new Unit(2, UnitTypes.Absolute) // Stop loss at 2*ATR
		);
	}

	private void ProcessCandle(ICandleMessage candle, decimal kalmanValue, decimal atrValue)
	{
		if (candle.State != CandleStates.Finished)
			return;

		if (!IsFormedAndOnlineAndAllowTrading())
			return;
		
		// Calculate trend direction
		var trend = candle.ClosePrice > kalmanValue ? 1 : -1;
		
		// Trading logic based on price position relative to Kalman filter
		if (trend > 0 && Position <= 0)
		{
			// Buy when price is above Kalman filter (uptrend)
			BuyMarket(Volume + Math.Abs(Position));
		}
		else if (trend < 0 && Position >= 0)
		{
			// Sell when price is below Kalman filter (downtrend)
			SellMarket(Volume + Math.Abs(Position));
		}
	}
}