oxictl 0.1.1

Pure Rust Real-Time Control Systems Framework
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140

use crate::core::scalar::PidScalar;

/// First-order IIR low-pass filter for the derivative term.
/// Implements the incomplete derivative: D(s) = Kd * s / (1 + tau_f * s)
/// Discretized using backward Euler: y[n] = alpha * y[n-1] + (1-alpha) * x[n]
/// where alpha = tau_f / (tau_f + dt)
#[derive(Debug, Clone, Copy)]
pub struct DerivativeFilter<S: PidScalar> {
    /// Filter time constant (seconds). Larger = more filtering.
    tau_f: S,
    /// Previous filtered output.
    prev_output: S,
    /// Whether the filter has been initialized.
    initialized: bool,
}

impl<S: PidScalar> DerivativeFilter<S> {
    /// Create a new derivative filter.
    /// `tau_f` is the filter time constant. Typical: tau_f = Kd / (N * Kp) where N = 8..20.
    pub fn new(tau_f: S) -> Self {
        Self {
            tau_f,
            prev_output: S::ZERO,
            initialized: false,
        }
    }

    /// Create from derivative gain and filter coefficient N.
    /// tau_f = Kd / (N * Kp). If Kp is zero, uses a default tau_f.
    pub fn from_gains(kd: S, kp: S, n: S) -> Self {
        let denominator = n * kp;
        let tau_f = if denominator.abs() > S::EPSILON {
            kd / denominator
        } else {
            kd / n
        };
        Self::new(tau_f)
    }

    /// Apply the filter to a raw derivative value.
    pub fn apply(&mut self, raw_derivative: S, dt: S) -> S {
        if !self.initialized {
            self.prev_output = raw_derivative;
            self.initialized = true;
            return raw_derivative;
        }

        if dt <= S::ZERO || self.tau_f <= S::ZERO {
            self.prev_output = raw_derivative;
            return raw_derivative;
        }

        let alpha = self.tau_f / (self.tau_f + dt);
        let filtered = alpha * self.prev_output + (S::ONE - alpha) * raw_derivative;
        self.prev_output = filtered;
        filtered
    }

    pub fn reset(&mut self) {
        self.prev_output = S::ZERO;
        self.initialized = false;
    }

    pub fn tau_f(&self) -> S {
        self.tau_f
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn first_value_passthrough() {
        let mut filt = DerivativeFilter::<f64>::new(0.1);
        let out = filt.apply(5.0, 0.01);
        assert_eq!(out, 5.0);
    }

    #[test]
    fn filter_attenuates_step_change() {
        let mut filt = DerivativeFilter::<f64>::new(0.1);
        filt.apply(0.0, 0.01);
        let out = filt.apply(10.0, 0.01);
        // alpha = 0.1/(0.1+0.01) ≈ 0.909
        // filtered = 0.909*0 + 0.091*10 ≈ 0.909
        assert!(out < 2.0, "Filter should attenuate: got {}", out);
        assert!(out > 0.0, "Filter should be positive: got {}", out);
    }

    #[test]
    fn filter_converges_to_constant() {
        let mut filt = DerivativeFilter::<f64>::new(0.01);
        filt.apply(0.0, 0.001);
        for _ in 0..1000 {
            filt.apply(5.0, 0.001);
        }
        let out = filt.apply(5.0, 0.001);
        assert!(
            (out - 5.0).abs() < 0.01,
            "Should converge to 5.0, got {}",
            out
        );
    }

    #[test]
    fn reset_clears_state() {
        let mut filt = DerivativeFilter::<f64>::new(0.1);
        filt.apply(100.0, 0.01);
        filt.reset();
        let out = filt.apply(0.0, 0.01);
        assert_eq!(out, 0.0);
    }

    #[test]
    fn from_gains_computes_tau() {
        let filt = DerivativeFilter::<f64>::from_gains(1.0, 2.0, 10.0);
        assert!((filt.tau_f() - 0.05).abs() < 1e-10);
    }

    #[test]
    fn noise_attenuation() {
        let mut filt = DerivativeFilter::<f64>::new(0.05);
        let dt = 0.001;
        // Feed alternating noise
        filt.apply(0.0, dt);
        let mut max_output = 0.0_f64;
        for i in 0..100 {
            let noisy = if i % 2 == 0 { 10.0 } else { -10.0 };
            let out = filt.apply(noisy, dt);
            max_output = max_output.max(out.abs());
        }
        // Filtered output should be much smaller than input amplitude
        assert!(
            max_output < 10.0,
            "Filter should attenuate noise: max={}",
            max_output
        );
    }
}