File size: 7,608 Bytes
f2a4894
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
import random
import numpy as np
import gradio as gr
import pandas as pd
from deap import base, creator, tools

# 全局变量存储优化状态
state = {
    "toolbox": None,
    "population": None,
    "current_idx": 0,
    "current_gen": 0,
    "history": [],
    "best_ind": None,
    "running": False,
    "params": {}
}


def create_individual(min_max):
    return [
        random.uniform(min_max["kp_min"], min_max["kp_max"]),
        random.uniform(min_max["ki_min"], min_max["ki_max"]),
        random.uniform(min_max["kd_min"], min_max["kd_max"])
    ]


def init_toolbox(min_max):
    if "FitnessMax" in creator.__dict__:
        del creator.FitnessMax
    if "Individual" in creator.__dict__:
        del creator.Individual

    creator.create("FitnessMax", base.Fitness, weights=(1.0,))
    creator.create("Individual", list, fitness=creator.FitnessMax)

    toolbox = base.Toolbox()
    toolbox.register("individual", tools.initIterate, creator.Individual,
                     lambda: create_individual(min_max))
    toolbox.register("population", tools.initRepeat, list, toolbox.individual)
    toolbox.register("mate", tools.cxBlend, alpha=0.5)
    toolbox.register("mutate", tools.mutGaussian, mu=0, sigma=0.1, indpb=0.2)
    toolbox.register("select", tools.selTournament, tournsize=3)
    return toolbox


def start_optimization(kp_min, kp_max, ki_min, ki_max, kd_min, kd_max,
                       pop_size, max_gen, cx_prob, mut_prob):
    state.update({
        "params": {
            "kp_range": (kp_min, kp_max),
            "ki_range": (ki_min, ki_max),
            "kd_range": (kd_min, kd_max),
            "pop_size": pop_size,
            "max_gen": max_gen,
            "cx_prob": cx_prob,
            "mut_prob": mut_prob
        },
        "current_gen": 0,
        "history": [],
        "best_ind": None,
        "running": True
    })

    min_max = {
        "kp_min": kp_min, "kp_max": kp_max,
        "ki_min": ki_min, "ki_max": ki_max,
        "kd_min": kd_min, "kd_max": kd_max
    }

    toolbox = init_toolbox(min_max)
    population = toolbox.population(n=pop_size)

    state.update({
        "toolbox": toolbox,
        "population": population,
        "current_idx": 0
    })

    first_ind = population[0]
    return {
        current_params: f"Kp: {first_ind[0]:.4f}, Ki: {first_ind[1]:.4f}, Kd: {first_ind[2]:.4f}",
        fitness_input: "",
        history_output: pd.DataFrame(columns=["Kp", "Ki", "Kd", "Fitness"]),
        best_output: "最佳参数:尚未找到"
    }


def submit_fitness(fitness):
    if not state["running"]:
        return {current_params: "优化未运行"}

    population = state["population"]
    idx = state["current_idx"]
    ind = population[idx]

    # 记录适应度
    ind.fitness.values = (float(fitness),)
    state["history"].append({
        "Kp": ind[0], "Ki": ind[1], "Kd": ind[2], "Fitness": float(fitness)
    })

    # 更新最佳个体
    if state["best_ind"] is None or ind.fitness > state["best_ind"].fitness:
        state["best_ind"] = ind

    # 移动到下一个个体
    state["current_idx"] += 1

    # 检查是否完成当前种群评估
    if state["current_idx"] >= len(population):
        evolve_population()
        state["current_idx"] = 0
        state["current_gen"] += 1

        # 检查终止条件
        if state["current_gen"] >= state["params"]["max_gen"]:
            state["running"] = False
            return {
                current_params: "优化完成!",
                best_output: format_best(state["best_ind"])
            }

    next_ind = population[state["current_idx"]]
    return {
        current_params: f"Kp: {next_ind[0]:.4f}, Ki: {next_ind[1]:.4f}, Kd: {next_ind[2]:.4f}",
        fitness_input: "",
        history_output: pd.DataFrame(state["history"]),
        best_output: format_best(state["best_ind"])
    }


def evolve_population():
    params = state["params"]
    toolbox = state["toolbox"]
    population = state["population"]

    # 选择
    offspring = toolbox.select(population, len(population))
    offspring = list(map(toolbox.clone, offspring))

    # 交叉
    for child1, child2 in zip(offspring[::2], offspring[1::2]):
        if random.random() < params["cx_prob"]:
            toolbox.mate(child1, child2)
            del child1.fitness.values
            del child2.fitness.values

    # 变异
    for mutant in offspring:
        if random.random() < params["mut_prob"]:
            toolbox.mutate(mutant)
            # 参数裁剪
            mutant[0] = np.clip(mutant[0], *params["kp_range"])
            mutant[1] = np.clip(mutant[1], *params["ki_range"])
            mutant[2] = np.clip(mutant[2], *params["kd_range"])
            del mutant.fitness.values

    state["population"] = offspring


def stop_optimization():
    state["running"] = False
    return {
        current_params: "优化已终止",
        fitness_input: ""
    }


def format_best(ind):
    if ind is None:
        return "最佳参数:尚未找到"
    return f"最佳参数:Kp={ind[0]:.4f}, Ki={ind[1]:.4f}, Kd={ind[2]:.4f} 适应度={ind.fitness.values[0]:.4f}"


with gr.Blocks(title="青云调参", css_paths="./style.css") as demo:
    gr.HTML("<h1 style='text-align: center;'>青云调参</h1><div style='text-align: center;'>适者存千代竞逐,精微处三昧调弦</div>")

    with gr.Row():
        with gr.Column(scale=1):
            with gr.Row():
                start_btn = gr.Button("开始优化", variant="primary")
                stop_btn = gr.Button("终止优化")

            with gr.Row():
                kp_min = gr.Number(label="Kp最小值", value=0.0)
                kp_max = gr.Number(label="Kp最大值", value=10.0)

            with gr.Row():
                ki_min = gr.Number(label="Ki最小值", value=0.0)
                ki_max = gr.Number(label="Ki最大值", value=10.0)

            with gr.Row():
                kd_min = gr.Number(label="Kd最小值", value=0.0)
                kd_max = gr.Number(label="Kd最大值", value=10.0)

            pop_size = gr.Slider(2, 50, value=5, step=1, label="种群大小")
            max_gen = gr.Slider(1, 50, value=10, step=1, label="迭代次数")
            cx_prob = gr.Slider(0.0, 1.0, value=0.7, step=0.05, label="交叉概率")
            mut_prob = gr.Slider(0.0, 1.0, value=0.2, step=0.05, label="变异概率")

        with gr.Column(scale=2):
            current_params = gr.Textbox(label="当前参数组合", interactive=False)
            with gr.Row():
                fitness_input = gr.Number(label="输入适应度值")
                submit_btn = gr.Button("提交适应度")

            gr.Markdown("## 最佳参数")
            best_output = gr.Textbox(label="当前最佳参数", interactive=False)

            gr.Markdown("## 优化历史")
            history_output = gr.Dataframe(
                headers=["Kp", "Ki", "Kd", "Fitness"],
                datatype=["number", "number", "number", "number"],
                interactive=False
            )

    start_btn.click(
        start_optimization,
        inputs=[kp_min, kp_max, ki_min, ki_max, kd_min, kd_max,
                pop_size, max_gen, cx_prob, mut_prob],
        outputs=[current_params, fitness_input, history_output, best_output]
    )

    submit_btn.click(
        submit_fitness,
        inputs=fitness_input,
        outputs=[current_params, fitness_input, history_output, best_output]
    )

    stop_btn.click(
        stop_optimization,
        outputs=[current_params, fitness_input]
    )

if __name__ == "__main__":
    demo.launch()