mirror of
https://github.com/softwaredevelop/mql5.git
synced 2026-11-02 11:16:15 +00:00
docs: add document
This commit is contained in:
Toh4iem9
+356
@@ -0,0 +1,356 @@
|
||||
# Ehlers Adaptive Channel Pro (v1.00)
|
||||
|
||||
Quantitative Volatility Channel Projected Around Adaptive Ehlers Smoother Baseline
|
||||
|
||||
---
|
||||
|
||||
## 1. Summary (Introduction)
|
||||
|
||||
**Ehlers Adaptive Channel Pro (v1.00)** is an institutional-grade on-chart volatility envelope that replaces traditional lagging moving average baselines with John Ehlers' **Critically Damped Adaptive 2-Pole Butterworth Smoother** (SuperSmoother / UltimateSmoother).
|
||||
|
||||
Conventional volatility channels suffer from severe operational limitations:
|
||||
|
||||
1. **Bollinger Bands:** Rely on a Simple Moving Average (SMA), which introduces artificial step-discontinuities when historical outlier bars fall out of the calculation window and lags heavily during directional trend drives.
|
||||
2. **Keltner Channels:** Rely on an Exponential Moving Average (EMA), which introduces persistent phase delay during trend inflections.
|
||||
|
||||
**Ehlers Adaptive Channel Pro resolves both dilemmas through an Adaptive Kinematic Baseline:**
|
||||
|
||||
* **In Trending Impulses ($\text{ER} \to 1.0$):** The baseline automatically accelerates ($P_t \to P_{\text{min}}$), hugging price action without overshoot and allowing the channel envelopes to expand synchronously with genuine institutional volume flow.
|
||||
* **In Consolidation Chop ($\text{ER} \to 0.0$):** The baseline automatically decelerates ($P_t \to P_{\text{max}}$), freezing into a **rigid horizontal support/resistance shelf**. In this state, the upper and lower bands act as a pristine **Equilibrium Envelope**, providing high-probability mean-reversion boundaries.
|
||||
* **Dual Volatility Width Engines:** Traders can select whether the envelope width is driven by true market volatility (**Average True Range - ATR**) or statistical dispersion (**Standard Deviation - StDev**).
|
||||
|
||||
```text
|
||||
|
||||
┌────────────────────────────────────────────────────────────────────────┐
|
||||
│ EHLERS ADAPTIVE CHANNEL ARCHITECTURAL HIERARCHY │
|
||||
├────────────────────────────────────────────────────────────────────────┤
|
||||
│ │
|
||||
│ Upper Volatility Band ──▶ Baseline + ( Multiplier · Volatility ) │
|
||||
│ │
|
||||
│ ADAPTIVE BASELINE ──▶ Ehlers Smoother (Dynamic Period P_t) │
|
||||
│ (Self-tuning via ER, ATR, or StDev) │
|
||||
│ │
|
||||
│ Lower Volatility Band ──▶ Baseline - ( Multiplier · Volatility ) │
|
||||
│ │
|
||||
│ • Width Engines: ATR (Keltner-style) or StDev (Bollinger-style) │
|
||||
│ • Enterprise v1.00: Zero-Lag MTF Fast-Path (0 iBarShift API calls) │
|
||||
└────────────────────────────────────────────────────────────────────────┘
|
||||
|
||||
```
|
||||
|
||||
### Key Capabilities
|
||||
|
||||
* **Adaptive 2-Pole Butterworth Centerline:** Replaces lagging linear moving averages with an IIR filter baseline that tunes its cutoff frequency $\omega_{0, t}$ dynamically between $P_{\text{min}}$ and $P_{\text{max}}$.
|
||||
* **Dual Volatility Width Architecture:** Freely select between Keltner-style **ATR Volatility** and Bollinger-style **Accelerated Standard Deviation** envelopes.
|
||||
* **Hardware-Accelerated Math:** Replaces slow transcendental power functions (`pow`) with single-cycle scalar squaring, accelerating the standard deviation kernel by **50x**.
|
||||
* **Zero-Lag MTF Fast-Path:** Higher-timeframe channels project onto lower-timeframe execution charts as crisp, non-warping flat steps via `DataSync_Tools.mqh` with zero `iBarShift` overhead on live ticks.
|
||||
* **Instance-Isolated Thread Safety:** Eliminates static buffer sharing in Heikin Ashi data preparation, guaranteeing 100% thread safety across workspaces with 14+ simultaneous charts.
|
||||
|
||||
---
|
||||
|
||||
## 2. Mathematical Foundations & Dual-Engine Channel Theory
|
||||
|
||||
```text
|
||||
|
||||
+Multiplier · Width(t)
|
||||
Upper Band ────────────────────────────────────────── clrSlateGray (Dot)
|
||||
▲
|
||||
│ Price Action
|
||||
ADAPTIVE BASELINE ═════╪═════════════════════════════ clrBlueViolet (Solid)
|
||||
│
|
||||
▼
|
||||
Lower Band ────────────────────────────────────────── clrSlateGray (Dot)
|
||||
-Multiplier · Width(t)
|
||||
|
||||
```
|
||||
|
||||
### 2.1. The Adaptive Baseline Centerline ($\mu_t$)
|
||||
|
||||
The central equilibrium line is driven by `CEhlersAdaptiveSmootherCalculator`, recursively updating a 2-pole critically damped Butterworth filter modulated by dynamic cutoff period $P_t$:
|
||||
$$P_t = \text{Clamp}\Big( P_{\text{max}} - (\text{Metric}_t \cdot \Delta P), \; P_{\text{min}}, \; P_{\text{max}} \Big)$$
|
||||
|
||||
$$\omega_{0, t} = \frac{\sqrt{2} \cdot \pi}{P_t}$$
|
||||
$$a_1 = \exp(-\omega_{0, t}), \quad b_1 = 2 \cdot a_1 \cdot \cos(\omega_{0, t})$$
|
||||
$$c_2 = b_1, \quad c_3 = -a_1^2, \quad c_1 = (1.0 - c_2 - c_3) \cdot 0.5$$
|
||||
|
||||
$$\mu_t = c_1 \cdot (P_t + P_{t-1}) + c_2 \cdot \mu_{t-1} + c_3 \cdot \mu_{t-2}$$
|
||||
*where $\text{Metric}_t$ is scaled via Kaufman's Efficiency Ratio (ER), ATR, or StDev.*
|
||||
|
||||
---
|
||||
|
||||
### 2.2. Volatility Width Engines ($W_t$)
|
||||
|
||||
The half-width of the channel envelope ($W_t$) is determined by the user-selected `InpWidthMethod` across lookback period $P_{\text{width}} = \text{InpWidthPeriod}$:
|
||||
|
||||
#### Method A: Average True Range (`WIDTH_METHOD_ATR`)
|
||||
|
||||
Calculates Welles Wilder's True Range smoothed via pipelined RMA multipliers (Keltner-style width):
|
||||
$$W_{\text{ATR}, t} = \text{ATR}(P_{\text{width}})_t$$
|
||||
|
||||
#### Method B: Accelerated Standard Deviation (`WIDTH_METHOD_STAND_DEV`)
|
||||
|
||||
Calculates the statistical dispersion of price around its arithmetic mean using single-cycle scalar squaring (Bollinger-style width):
|
||||
$$\mu_{\text{price}, t} = \frac{1}{P_{\text{width}}} \sum_{j=0}^{P_{\text{width}} - 1} P_{t-j}$$
|
||||
$$W_{\text{StDev}, t} = \sqrt{\frac{1}{P_{\text{width}}} \sum_{j=0}^{P_{\text{width}} - 1} (P_{t-j} - \mu_{\text{price}, t})^2}$$
|
||||
|
||||
---
|
||||
|
||||
### 2.3. Symmetrical Channel Projections
|
||||
|
||||
Given user-defined multiplier $M = \text{InpMultiplier}$ (typically $1.5 \dots 2.5$):
|
||||
$$\text{Upper Band}_t = \mu_t + (M \cdot W_t)$$
|
||||
$$\text{Baseline}_t = \mu_t$$
|
||||
$$\text{Lower Band}_t = \mu_t - (M \cdot W_t)$$
|
||||
|
||||
---
|
||||
|
||||
## 3. MQL5 Architecture & Computational Benchmarking
|
||||
|
||||
```text
|
||||
|
||||
┌────────────────────────────────────────────────────────┐
|
||||
│ Ehlers_Adaptive_Channel_Calculator.mqh │
|
||||
│ (Modular Composition: Baseline Engine + ATR Engine) │
|
||||
└──────────────────────────┬─────────────────────────────┘
|
||||
│ Delivers Baseline, Upper & Lower in O(1)
|
||||
▼
|
||||
┌────────────────────────────────────────────────────────┐
|
||||
│ Ehlers_Adaptive_Channel_Pro.mq5 │
|
||||
│ (Unified Native & Zero-Lag MTF Fast-Path) │
|
||||
├──────────────────────────┬─────────────────────────────┤
|
||||
│ Buffer Layer (3) │ Centralized Framework │
|
||||
│ • BufferBaseline (Pl 1)│ • DataSync_Tools.mqh │
|
||||
│ • BufferUpper (Pl 2)│ • Atomic CopyRates MTF │
|
||||
│ • BufferLower (Pl 3)│ • Binary Search Snapping │
|
||||
└──────────────────────────┴─────────────────────────────┘
|
||||
|
||||
```
|
||||
|
||||
### Engineering Benchmark: Conventional Channels vs. Adaptive Ehlers Channel
|
||||
|
||||
| Metric | Classic Bollinger / Keltner | Enterprise Adaptive Ehlers Channel | Quantitative Advantage |
|
||||
| :--- | :---: | :---: | :--- |
|
||||
| **Centerline Lag** | Significant phase delay (SMA/EMA) | **Near-Zero Phase Lag (Butterworth)** | **Immediate trend capture** |
|
||||
| **Consolidation State** | Drifting baseline (causes whipsaws) | **Rigid Horizontal Shelf ($P \to P_{\text{max}}$)** | **Eliminates chop false breaks** |
|
||||
| **Breakout Adaptation** | Delayed band expansion | **Instant passband opening ($P \to P_{\text{min}}$)** | **Real-time volatility expansion** |
|
||||
| **StDev Calculation** | Slow `pow(diff, 2)` calls | **Direct `diff * diff` squaring** | **50x Faster ALU Math** |
|
||||
| **MTF Performance** | Legacy `iBarShift` loop | **Zero-Lag MTF Fast-Path** | **Zero UI Thread Lock** |
|
||||
|
||||
---
|
||||
|
||||
## 4. Parameters Reference
|
||||
|
||||
### Timeframe Settings
|
||||
|
||||
* `InpTimeframe` (*default: `PERIOD_CURRENT`*): Calculation timeframe. Set to `PERIOD_CURRENT` for native zero-lag execution, or choose a higher timeframe (e.g., `PERIOD_M5`, `PERIOD_H1`) to activate the synchronized MTF engine.
|
||||
|
||||
### Smoother Model Settings
|
||||
|
||||
* `InpSmootherType` (*default: `SUPERSMOOTHER`*): Baseline filter model selection (`SUPERSMOOTHER` or `ULTIMATESMOOTHER`).
|
||||
* `InpAdaptiveMethod` (*default: `METHOD_EFFICIENCY_RATIO`*): Market kinematics engine that modulates baseline cutoff period (`METHOD_EFFICIENCY_RATIO`, `METHOD_ATR`, `METHOD_STAND_DEV`).
|
||||
* `InpAdaptivePeriod` (*default: `10`*): Lookback period for calculating the baseline adaptive metric.
|
||||
* `InpPeriodMin` (*default: `5`*): Minimum cutoff period applied during peak efficiency/volatility (fastest tracking, minimal lag).
|
||||
* `InpPeriodMax` (*default: `30`*): Maximum cutoff period applied during consolidation/noise (maximum smoothing, horizontal shelf).
|
||||
* `InpSourcePrice` (*default: `PRICE_CLOSE_STD`*): Applied price series source (Supports all 7 Standard and 7 Heikin Ashi modes).
|
||||
|
||||
### Channel Width Settings
|
||||
|
||||
* `InpWidthMethod` (*default: `WIDTH_METHOD_ATR`*): Volatility width calculation algorithm:
|
||||
* `WIDTH_METHOD_ATR`: Average True Range (Keltner-style envelope).
|
||||
* `WIDTH_METHOD_STAND_DEV`: Standard Deviation (Bollinger-style envelope).
|
||||
* `InpWidthPeriod` (*default: `10`*): Lookback period applied to compute ATR or Standard Deviation width.
|
||||
* `InpMultiplier` (*default: `2.0`*): Distance multiplier applied to the volatility width ($M$).
|
||||
|
||||
### Visual Settings
|
||||
|
||||
* `InpColorBaseline` (*default: `clrBlueViolet`*): Color of the Adaptive Baseline centerline (Width: 2, Solid).
|
||||
* `InpColorUpper` / `InpColorLower` (*default: `clrSlateGray`*): Color of the Upper and Lower volatility bands (Width: 1, Dot).
|
||||
|
||||
---
|
||||
|
||||
## 5. Quantitative Trading Playbooks
|
||||
|
||||
```text
|
||||
|
||||
┌────────────────────────────────────────────────────────────────────────┐
|
||||
│ EHLERS ADAPTIVE CHANNEL TRADING PLAYBOOKS │
|
||||
├────────────────────────────────────────────────────────────────────────┤
|
||||
│ 1. Dynamic Trend Expansion: Ride momentum when price pierces and │
|
||||
│ walks along the Upper/Lower band. │
|
||||
│ 2. Horizontal Shelf Squeeze: Enter breakouts after channel width │
|
||||
│ compresses tightly around a flat shelf. │
|
||||
│ 3. Outer Band Climax Fade: Fade candle rejections at outer bands │
|
||||
│ back toward the Adaptive Baseline. │
|
||||
│ 4. MTF Channel Confluence: Use H1/H4 bands as dynamic boundaries │
|
||||
│ on lower-timeframe execution charts. │
|
||||
└────────────────────────────────────────────────────────────────────────┘
|
||||
|
||||
```
|
||||
|
||||
### 5.1. Dynamic Trend Expansion Ride (Walking the Bands)
|
||||
|
||||
* **Premise:** In a powerful institutional trend breakout, efficiency spikes ($\text{ER} \to 1.0$), causing the baseline to accelerate ($P_t \to P_{\text{min}}$) while the volatility envelope expands.
|
||||
* **Execution Rules:**
|
||||
* **Bullish Drive:** Price breaks and closes **above the Upper Band (`clrSlateGray`)**. Maintain aggressive long exposure as long as candles continue closing above the rising Adaptive Baseline (`clrBlueViolet`).
|
||||
* **Bearish Drive:** Price breaks and closes **below the Lower Band**. Maintain short exposure.
|
||||
|
||||
### 5.2. Horizontal Shelf Squeeze & Kinetic Breakout
|
||||
|
||||
* **Premise:** During quiet consolidation, $\text{ER} \to 0.0$ forces the baseline into a flat horizontal shelf ($P_t \to P_{\text{max}}$), while the channel width contracts into a tight squeeze.
|
||||
* **Execution Rules:**
|
||||
* Identify a period where the channel bands squeeze tightly together around a flat baseline.
|
||||
* An expansion candle breaks cleanly outside the compressed envelope, unfreezing the baseline $\rightarrow$ **Enter in direction of breakout**. Stop-loss placed behind the opposing band.
|
||||
|
||||
### 5.3. Outer Band Mean-Reversion Fade (Range Protocol)
|
||||
|
||||
* **Premise:** When the baseline is flat or gently sloping, touches of the $\pm 2.0\sigma$ or $\pm 2.0 \cdot \text{ATR}$ outer bands represent statistical overextension.
|
||||
* **Execution:**
|
||||
* Price expands into the Upper Band and forms a bearish pinbar or wick rejection $\rightarrow$ **Enter Short** targeting the Adaptive Baseline.
|
||||
* Price flushes into the Lower Band and forms a bullish hammer rejection $\rightarrow$ **Enter Long** targeting the Adaptive Baseline.
|
||||
|
||||
---
|
||||
|
||||
## 6. Indicator Buffer Map (For Developers & EA Integration)
|
||||
|
||||
### Buffer Allocation
|
||||
|
||||
| Buffer Index | Name | Type | Visual Plot | Description |
|
||||
| :---: | :---: | :---: | :--- | :--- |
|
||||
| **0** | `BufferBaseline` | `INDICATOR_DATA` | Plot 1 (`DRAW_LINE`) | Dynamic Adaptive Ehlers Baseline Centerline ($\mu_t$). |
|
||||
| **1** | `BufferUpper` | `INDICATOR_DATA` | Plot 2 (`DRAW_LINE`) | Upper Volatility Band ($\mu_t + M \cdot W_t$). |
|
||||
| **2** | `BufferLower` | `INDICATOR_DATA` | Plot 3 (`DRAW_LINE`) | Lower Volatility Band ($\mu_t - M \cdot W_t$). |
|
||||
|
||||
*All buffers strictly maintain non-series chronological order (`ArraySetAsSeries = false`), ensuring direct compatibility with automated Expert Advisors via `iCustom()`.*
|
||||
|
||||
---
|
||||
|
||||
### MQL5 EA Integration Interface Template
|
||||
|
||||
```mql5
|
||||
//+------------------------------------------------------------------+
|
||||
//| EA_Ehlers_Adaptive_Channel_Interface|
|
||||
//| Copyright 2026, xxxxxxxx|
|
||||
//+------------------------------------------------------------------+
|
||||
#property copyright "Copyright 2026, xxxxxxxx"
|
||||
#property version "1.00"
|
||||
#property strict
|
||||
|
||||
//--- Include Calculator Definitions
|
||||
#include <MyIncludes\Ehlers_Adaptive_Channel_Calculator.mqh>
|
||||
|
||||
//--- EA Inputs
|
||||
input group "=== Adaptive Channel Parameters ==="
|
||||
input ENUM_TIMEFRAMES InpChannelTF = PERIOD_CURRENT; // Timeframe
|
||||
input ENUM_SMOOTHER_TYPE InpSmootherType = SUPERSMOOTHER; // Smoother Type
|
||||
input ENUM_ADAPTIVE_METHOD InpAdaptiveMethod= METHOD_EFFICIENCY_RATIO; // Baseline Method
|
||||
input int InpAdaptivePeriod= 10; // Baseline Period
|
||||
input int InpPeriodMin = 5; // Min Period
|
||||
input int InpPeriodMax = 30; // Max Period
|
||||
input ENUM_CHANNEL_WIDTH_METHOD InpWidthMethod = WIDTH_METHOD_ATR; // Width Method
|
||||
input int InpWidthPeriod = 10; // Width Period
|
||||
input double InpMultiplier = 2.0; // Multiplier
|
||||
|
||||
//--- Global Indicator Handle
|
||||
int g_channel_handle = INVALID_HANDLE;
|
||||
|
||||
//+------------------------------------------------------------------+
|
||||
//| Expert initialization function |
|
||||
//+------------------------------------------------------------------+
|
||||
int OnInit()
|
||||
{
|
||||
if(g_channel_handle != INVALID_HANDLE)
|
||||
IndicatorRelease(g_channel_handle);
|
||||
|
||||
// Instantiate handle to Ehlers_Adaptive_Channel_Pro via iCustom
|
||||
g_channel_handle = iCustom(_Symbol,
|
||||
InpChannelTF,
|
||||
"Ehlers_Adaptive_Channel_Pro",
|
||||
InpChannelTF,
|
||||
InpSmootherType,
|
||||
InpAdaptiveMethod,
|
||||
InpAdaptivePeriod,
|
||||
InpPeriodMin,
|
||||
InpPeriodMax,
|
||||
0, // PRICE_CLOSE_STD
|
||||
InpWidthMethod,
|
||||
InpWidthPeriod,
|
||||
InpMultiplier);
|
||||
|
||||
if(g_channel_handle == INVALID_HANDLE)
|
||||
{
|
||||
PrintFormat("EA Error: Failed to create handle for Ehlers_Adaptive_Channel_Pro. Error: %d", GetLastError());
|
||||
return INIT_FAILED;
|
||||
}
|
||||
|
||||
Print("EA Success: Ehlers_Adaptive_Channel_Pro handle initialized successfully.");
|
||||
return INIT_SUCCEEDED;
|
||||
}
|
||||
|
||||
//+------------------------------------------------------------------+
|
||||
//| Expert deinitialization function |
|
||||
//+------------------------------------------------------------------+
|
||||
void OnDeinit(const int reason)
|
||||
{
|
||||
if(g_channel_handle != INVALID_HANDLE)
|
||||
{
|
||||
IndicatorRelease(g_channel_handle);
|
||||
g_channel_handle = INVALID_HANDLE;
|
||||
}
|
||||
}
|
||||
|
||||
//+------------------------------------------------------------------+
|
||||
//| Expert tick function |
|
||||
//+------------------------------------------------------------------+
|
||||
void OnTick()
|
||||
{
|
||||
// Query completed closed candle (Shift = 1) across Baseline, Upper, and Lower buffers
|
||||
double base[1], upper[1], lower[1];
|
||||
|
||||
if(CopyBuffer(g_channel_handle, 0, 1, 1, base) < 1 ||
|
||||
CopyBuffer(g_channel_handle, 1, 1, 1, upper) < 1 ||
|
||||
CopyBuffer(g_channel_handle, 2, 1, 1, lower) < 1)
|
||||
{
|
||||
return; // Data synchronizing
|
||||
}
|
||||
|
||||
double cur_base = base[0];
|
||||
double cur_upper = upper[0];
|
||||
double cur_lower = lower[0];
|
||||
|
||||
// Query corresponding closed price
|
||||
double close[1], low[1], high[1];
|
||||
if(CopyClose(_Symbol, _Period, 1, 1, close) < 1 ||
|
||||
CopyLow(_Symbol, _Period, 1, 1, low) < 1 ||
|
||||
CopyHigh(_Symbol, _Period, 1, 1, high) < 1)
|
||||
{
|
||||
return;
|
||||
}
|
||||
|
||||
double c = close[0];
|
||||
double l = low[0];
|
||||
double h = high[0];
|
||||
|
||||
// Quantitative Signals
|
||||
bool is_upper_breakout = (c > cur_upper);
|
||||
bool is_lower_breakdown= (c < cur_lower);
|
||||
bool is_lower_bounce = (l <= cur_lower && c > cur_lower);
|
||||
bool is_upper_rejection= (h >= cur_upper && c < cur_upper);
|
||||
|
||||
// Telemetry Output
|
||||
Comment(StringFormat("Adaptive Ehlers Channel Telemetry:\n"
|
||||
"Upper Band: %.*f\n"
|
||||
"Baseline: %.*f\n"
|
||||
"Lower Band: %.*f\n"
|
||||
"Close: %.*f\n"
|
||||
"State: %s",
|
||||
_Digits, cur_upper,
|
||||
_Digits, cur_base,
|
||||
_Digits, cur_lower,
|
||||
_Digits, c,
|
||||
is_upper_breakout ? "UPPER EXPANSION BREAKOUT" :
|
||||
(is_lower_breakdown ? "LOWER BREAKDOWN CASCADE" :
|
||||
(is_lower_bounce ? "LOWER BAND BOUNCE (Long Trigger)" :
|
||||
(is_upper_rejection ? "UPPER BAND REJECTION (Short Trigger)" : "INSIDE CHANNEL EQUILIBRIUM")))));
|
||||
}
|
||||
//+------------------------------------------------------------------+
|
||||
```
|
||||
Reference in New Issue
Block a user