在scipy.optimize.curve_fit中有没有优化p0的好方法？

import numpy as np
from scipy.optimize import curve_fit
import matplotlib.pyplot as plt

# 定义一个简单的非线性函数
def func(x, a, b, c):
    return a * np.exp(-b * x) + c

# 生成一些带有噪声的数据
xdata = np.linspace(0, 4, 50)
ydata = func(xdata, 2.5, 1.3, 0.5) + 0.2 * np.random.normal(size=50)

# 尝试不同的初始猜测值
p0_guesses = [(1, 1, 1), (2, 2, 2), (3, 3, 3)]

best_popt = None
best_pcov = None
best_loss = np.inf

for p0 in p0_guesses:
    try:
        popt, pcov = curve_fit(func, xdata, ydata, p0=p0)
        loss = np.sum((func(xdata, *popt) - ydata) ** 2)
        if loss < best_loss:
            best_popt = popt
            best_pcov = pcov
            best_loss = loss
    except RuntimeError:
        # 如果curve_fit无法收敛，则跳过当前的p0
        continue

print(f"Best fit parameters: {best_popt}")

# 绘制结果
plt.scatter(xdata, ydata, label='Data')
plt.plot(xdata, func(xdata, *best_popt), 'r-', label='Fit')
plt.legend()
plt.show()