algorithm_platform/models_processing/model_koi/tf_bp.py

#!/usr/bin/env python
# -*- coding:utf-8 -*-
# @FileName  :nn_bp.py
# @Time      :2025/2/12 10:41
# @Author    :David
# @Company: shenyang JY


from tensorflow.keras.layers import Input, Dense, LSTM, concatenate, Conv1D, Conv2D, MaxPooling1D, Reshape, Flatten
from tensorflow.keras.models import Model, load_model
from tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping, TensorBoard, ReduceLROnPlateau
from tensorflow.keras import optimizers, regularizers
import numpy as np
from common.database_dml import *
from threading import Lock
model_lock = Lock()

class BPHandler(object):
    def __init__(self, logger):
        self.logger = logger
        self.model = None

    def get_model(self, args):
        """
        单例模式+线程锁，防止在异步加载时引发线程安全
        """
        try:
            with model_lock:
                # NPHandler.model = NPHandler.get_keras_model(opt)
                self.model = get_h5_model_from_mongo(args)
        except Exception as e:
            self.logger.info("加载模型权重失败:{}".format(e.args))

    @staticmethod
    def get_keras_model(opt):
        # db_loss = NorthEastLoss(opt)
        # south_loss = SouthLoss(opt)
        from models_processing.losses.loss_cdq import rmse
        l1_reg = regularizers.l1(opt.Model['lambda_value_1'])
        l2_reg = regularizers.l2(opt.Model['lambda_value_2'])
        nwp_input = Input(shape=(opt.Model['time_step'], opt.Model['input_size']), name='nwp')

        con1 = Conv1D(filters=64, kernel_size=1, strides=1, padding='valid', activation='relu', kernel_regularizer=l2_reg)(nwp_input)
        d1 = Dense(32, activation='relu', name='d1', kernel_regularizer=l1_reg)(con1)
        nwp = Dense(8, activation='relu', name='d2', kernel_regularizer=l1_reg)(d1)

        output = Dense(1, name='d5')(nwp)
        output_f = Flatten()(output)
        model = Model(nwp_input, output_f)
        adam = optimizers.Adam(learning_rate=opt.Model['learning_rate'], beta_1=0.9, beta_2=0.999, epsilon=1e-7, amsgrad=True)
        reduce_lr = ReduceLROnPlateau(monitor='val_loss', factor=0.01, patience=5, verbose=1)
        model.compile(loss=rmse, optimizer=adam)
        return model

    def train_init(self, opt):
        try:
            if opt.Model['add_train']:
                # 进行加强训练，支持修模
                base_train_model = get_h5_model_from_mongo(vars(opt))
                base_train_model.summary()
                self.logger.info("已加载加强训练基础模型")
            else:
                base_train_model = self.get_keras_model(opt)
            return base_train_model
        except Exception as e:
            self.logger.info("加强训练加载模型权重失败:{}".format(e.args))

    def training(self, opt, train_and_valid_data):
        model = self.train_init(opt)
        # tf.reset_default_graph() # 清除默认图
        train_x, train_y, valid_x, valid_y = train_and_valid_data
        print("----------", np.array(train_x[0]).shape)
        print("++++++++++", np.array(train_x[1]).shape)
        model.summary()
        check_point = ModelCheckpoint(filepath='./var/' + 'fmi.h5', monitor='val_loss',  save_best_only=True, mode='auto')
        early_stop = EarlyStopping(monitor='val_loss', patience=opt.Model['patience'], mode='auto')
        history = model.fit(train_x, train_y, batch_size=opt.Model['batch_size'], epochs=opt.Model['epoch'], verbose=2,  validation_data=(valid_x, valid_y), callbacks=[check_point, early_stop], shuffle=False)
        loss = np.round(history.history['loss'], decimals=5)
        val_loss = np.round(history.history['val_loss'], decimals=5)
        self.logger.info("-----模型训练经过{}轮迭代-----".format(len(loss)))
        self.logger.info("训练集损失函数为：{}".format(loss))
        self.logger.info("验证集损失函数为：{}".format(val_loss))
        return model

    def predict(self, test_x, batch_size=1):
        result = self.model.predict(test_x, batch_size=batch_size)
        self.logger.info("执行预测方法")
        return result


if __name__ == "__main__":
    run_code = 0