config_b_DNN_engagement_prediction.py

import numpy as np
import pandas as pd
import tensorflow as tf


## Data Preprocessing
# Importing the dataset
dataset = pd.read_excel("prediect_result_cofig_b.xlsx")
X = dataset.iloc[1:, :3].values
Y = dataset.iloc[1:, 4].values
Y = np.where(Y == 0, 0, 1)

# Splitting the dataset into the Training set and Test set
from sklearn.model_selection import train_test_split

X_train, X_test, Y_train, Y_test = train_test_split(X, Y, test_size=0.2, random_state=0)

X_train = np.array(X_train, dtype=np.float32)
Y_train = np.array(Y_train, dtype=np.float32)

# # Feature Scaling
import joblib

scx = joblib.load("scaler.pkl")
X_train = scx.fit_transform(X_train)
X_test = scx.transform(X_test)


# Loading the model from the file
loaded_model = tf.keras.models.load_model("config_b_DNN_model.h5")

# Making predictions with the loaded model
Y_pred = loaded_model.predict(X_test)
Y_pred = np.round(Y_pred)


# Visualising the Test set results in black and white
import matplotlib.pyplot as plt
from matplotlib.colors import ListedColormap
import numpy as np

X_set, y_set = scx.inverse_transform(X_test), Y_test

X1, X2 = np.meshgrid(
    np.arange(
        start=X_set[:, 0].min() - 45.0, stop=X_set[:, 0].max() + 45.0, step=0.0450
    ),
    np.arange(
        start=X_set[:, 1].min() - 0.01, stop=X_set[:, 1].max() + 0.01, step=0.0005
    ),
)


plt.contourf(
    X1,
    X2,
    loaded_model.predict(
        scx.transform(np.array([X1.ravel(), X2.ravel(), np.ones_like(X1.ravel())]).T)
    ).reshape(X1.shape),
    alpha=0.75,
    cmap=ListedColormap(("white", "black")),
)

plt.xlim(X1.min(), X1.max())
plt.ylim(X2.min(), X2.max())


# Set the font to Times New Roman
from matplotlib import rcParams

rcParams["font.family"] = "serif"
rcParams["font.serif"] = ["Times New Roman"]

for i, label in enumerate(["Not engaged", "Engaged"]):
    plt.scatter(
        X_set[y_set == i, 0],
        X_set[y_set == i, 1],
        c=ListedColormap(("white", "black"))(i),
        edgecolor="black",
        label=label,
        s=10,
    )


plt.title(
    r"Deep Learning Based Centrifugal Clutch Engagement Prediction",
    fontdict={"family": "Times New Roman", "size": 14, "weight": "bold"},
)
plt.xlabel(r"$F_{\mathrm{preload}}$ ${(N)}$")
plt.ylabel(r"$m_{\mathrm{shoe}}$ ${(g)}$")

plt.xticks(fontsize=9.5)
plt.yticks(fontsize=9.5)

plt.legend(loc="lower right", labelspacing=0.1, handletextpad=0.2, framealpha=1.0)

plt.show()